article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
blazars are the most powerful active galactic nuclei ( agn ) observed and are remarkable for their variability .
their most prominent characteristic is the relativistic jets which are aligned with the line of sight to the observer . in this preferred direction
the blazar can be very bright .
although they are detectable over a wide range of frequencies , their alignment does not permit spatial resolution and makes the detection of optical spectral lines very difficult .
their broadband emission is clearly nonthermal and the nature of the progenitor particles ( electrons or protons ) is uncertain . in principle , there is much observational evidence which should make their understanding straightforward , but in practice there is a wealth of often contradictory observations which demonstrate that the underlying mechanisms are complex and often ambiguous . while their variability makes these agn inherently interesting , observations ( preferably simultaneous ) over a long time interval at a variety of wavelengths
are required to draw any firm conclusions . the spectral energy distribution ( sed ) of blazars is generally represented by a double - peaked structure , with the lower - energy peak modeled by a synchrotron emission mechanism in which the synchrotron photons arise from electrons in the relativistic beam ; the compton scattering of these electrons on the low - energy photons ( which may be either of synchrotron origin or external from some other mechanism ) results in high - energy gamma - ray emission @xcite .
generally , the former simple model ( i.e. , self - compton ) is preferred , but some sources appear to require external photons @xcite . in another class of model ,
hadronic interactions are invoked to explain the higher - energy peak @xcite .
markarian 421 ( mrk421 ) was one of 70 agn reported in the third egret catalog of 100 mev sources @xcite .
it was remarkable in that it was both the weakest and the closest ( redshift @xmath1 ) of these agn .
its variability was not particularly noteworthy in the discovery gamma - ray observation .
subsequently it was determined that , unlike the other agn in the egret catalog , it should be classified as a hbl ( high - frequency - peaked bl lacertae object ) in which the synchrotron and compton peaks in its spectral energy distribution were displaced some decades to higher frequency relative to the norm of egret - detected agn .
this made it a prime candidate for tev emission .
it was on a short list of egret - detected agn selected for observation at tev energies with the whipple 10 m telescope @xcite .
it was the only one on the list which gave evidence of a signal , and the subsequent paper @xcite announced the first unambiguous detection of an extragalactic source at tev energies . since that time
, mrk421 has been observed every season at the whipple observatory .
although more than 50 agn have now been detected at tev energies @xcite , mrk421 is still , on average , the strongest agn source of tev gamma rays in the northern hemisphere .
because of its proximity and high degree of variability , it was considered a good candidate for the detailed study of an agn that was a tev gamma - ray emitter . during the large flare in 2001 ,
it was observed to reach a peak emission of 13 times the level of the crab nebula ( the brightest known steady tev source ) over a four - minute integration time and to show other rapid variations @xcite .
atmospheric cherenkov telescopes are particularly useful for the study of hbls such as mrk421 because of their high sensitivity to rapid variations .
the whipple telescope has detected flux variability with doubling times as short as 15 minutes @xcite ; there is some correlation between emission at x - ray kev energies and gamma - ray tev energies @xcite .
rapid variations in tev emission have also been observed in other agn @xcite .
it has been shown that the energy spectrum of mrk421 hardens with increasing intensity @xcite . in this paper
, we summarize intensive observations of mrk421 with a single gamma - ray telescope over a long period .
the principal telescope used in this study , the 10 m reflector and imaging camera at the whipple observatory , is described in section 2 .
section 3 describes the database of tev observations taken over the 14-year epoch .
correlations with simultaneous x - ray observations are presented in section 4 .
the level of variability in the tev signal is explored in section 5 .
section 6 explores those intervals in which the tev signal was particularly strong and the source may be said to have been flaring .
the principal data reported here were taken at tev energies with the whipple 10 m atmospheric cherenkov telescope and imaging camera over a 14-year period ( december 1995 may 2009 ) .
although observations were made both before and after this period , this time interval was chosen as it represents the best period of uniform operation and performance of the telescope .
the whipple photomultiplier ( pmt ) camera evolved during this time @xcite . from 2001 to 2009 , the camera consisted of a hexagonal array of 379 pmts of diameter 1.2 cm , giving a total field of view of @xmath2 .
winston light cones in front of the pmts minimized the light loss and gave a relatively uniform sensitivity across the face of the camera .
the camera was triggered when the light level in at least three tubes exceeded a preset threshold . before 2001 ,
the camera had pmts of 2.5 cm diameter and had a larger field of view .
the whipple observations were supplemented by observations with the veritas ( very energetic radiation imaging telescope array system ) observatory , located at the basecamp of the fred lawrence whipple observatory in southern arizona , usa ; this was completed in june 2007 .
the observatory consists of four 12m - diameter imaging atmospheric cherenkov telescopes , with a typical baseline between telescopes of @xmath3100 m @xcite .
each telescope has a 499-pmt camera , spanning a field of view of @xmath4 .
the whipple telescope observations were taken in intervals of 28 sidereal minutes during which the telescope was directed at the source .
this constituted an on run . in order to estimate the background , the on run was sometimes followed ( or preceded ) by an off run during which the telescope tracked a point 30 minutes later ( earlier ) in right ascension but at the same declination , so that telescope traced out the same path in azimuth and elevation as the on run .
often no off runs were taken , so that the source was tracked continuously ( tracking mode ) and the background determined from the known image parameter distributions in the absence of a source .
observations were taken on clear moonless nights .
for much of this time , mrk421 was the prime target , and it was observed whenever it was at an elevation greater than @xmath5 . where possible , one or two on / off pairs were also taken on the crab nebula each night for calibration purposes .
the rest of the darktime was used to observe other candidate sources ( mostly agn ) or gamma - ray bursts .
the cherenkov images of air showers recorded with the telescope were characterized by the moments of the angular distribution of light as determined by the output of the pmts .
the division of the data into candidate gamma - ray images and background images was made on the basis of selection criteria optimized by monte carlo simulations of gamma - ray and hadron shower images and checked ( and again optimized ) on known sources such as the crab nebula @xcite .
on - line analysis was performed at the conclusion of each data run so that the state of the source was known promptly and the observing program adjusted if necessary .
only data taken on clear nights ( judged by fluctuations in the background trigger rate and by the observers comments ) were used in the final database . at the beginning of each observing season , the trigger threshold was raised or lowered , if required , to compensate for changes in the telescope and in the camera ; these changes included replacement of pmts , recoating of mirrors , upgrades to trigger electronics , optimisation of mirror - facet alignments , and improvements in telescope pointing corrections .
the adjustment to the trigger level from one season to the next was never more than @xmath6% , and the threshold was always maintained well above the noise trigger level . in order to analyze all the mrk421 data in a uniform manner , the average crab nebula flux for each observing season was used for calibration . although the mev - gev flux from the crab nebula has recently been shown to exhibit short - term flares @xcite , they are infrequent ( @xmath72 per year ) and their duty cycle is short ( @xmath310 days ) ; a search of the whipple database @xcite did not show any evidence for such flares over a ten - year epoch .
long - term secular variations such as those seen in x - ray experiments @xcite have not been seen at tev gamma - ray energies , so that the crab nebula can still be regarded as a standard candle in this energy regime .
on - source observations of mrk421 ( on runs and tracking runs ) were chosen and selection criteria as described above applied to identify gamma - ray candidate events .
estimation of the background was based on the distribution of the angle @xmath8 between the major axis of the event image and the line from camera centre to image centroid .
the region @xmath9 was taken as the signal region and the region @xmath10 as control region .
the background in the signal region was estimated by multiplying the number of events in the control region by a `` tracking ratio '' .
the tracking ratio @xmath11 for each season was determined from a set of `` darkfield '' runs ( i.e. , observing runs with no known source in the field of view ) as the number of size- and shape - selected events in the signal region divided by the corresponding number of events in the control region @xcite .
usually the off runs from on / off observations of mrk421 were used as darkfield runs , but where these were insufficient they were supplemented by off data on other sources . for an on - source run of duration @xmath12 , the gamma - ray rate
was then found as @xmath13 , where @xmath14 and @xmath15 are the number of size- and shape - selected events in the signal and control regions , respectively .
the error on the rate was calculated as @xmath16 , where @xmath17 is the error on the tracking ratio ( typically @xmath313% ) .
the significance of the detection was taken as @xmath18 ; this formulation was used in preference to that of li & ma @xcite in order to take account of the error on the tracking ratio . since gamma - ray observations
could only be taken with the whipple 10 m telescope on moonless nights , there were about eight nights around the time of full moon of each lunar cycle when no gamma - ray observations could be made .
the periods ( @xmath321 days each ) during which the gamma - ray data were taken are referred to as `` darkruns '' . typically , each observing season for mrk421 at the whipple observatory spans between 6 and 8 darkruns , from november december to may june .
the gamma - ray data for mrk421 between december 1995 and may 2009 , in runs of ( generally ) 28-minute duration , comprise the database for this work ; in all cases , the measured flux is given in crab units based on the crab signal recorded that season .
these mrk421 runs were combined ( weighted average ) to give the mean flux over larger time intervals : nightly ( individual nights of 16 hour duration ) , monthly ( entire dark runs ) and yearly .
the peak energy response of the telescope during these observations was 400 gev with a 20% uncertainty . in this work , the assumption is made that the crab nebula flux at tev energies is constant over the season .
the typical nightly rate from the crab nebula as measured by the whipple telescope for one season is shown in figure [ fig1 ] . although the energy spectrum of mrk421 has been observed to harden when the flux level is high @xcite , the spectral shape in the energy range 400 gev to 4 tev is typically similar to that of the crab nebula , so that for most nights ( when mrk421 is not flaring strongly ) the rates expressed in crab units are insensitive to small changes in peak energy response . some results from this data set
have been published previously @xcite .
several of these reports were motivated by the observation of exceptional activity ( may 1996 @xcite ; february 2001 @xcite ; may 2008 @xcite ) and the reports included light curves on short timescales and spectral analysis .
these flares are discussed in section 6 .
the mrk421 observations reported here represent a total exposure time of 878.4 hours over 14 years .
the nightly gamma - ray rate from mrk421 for the typical 20072008 observing season , expressed in crab units , is shown in figure [ fig2 ] ; the horizontal line indicates the average rate for this season .
mrk421 varies on many time scales .
the mean gamma - ray rate from mrk421 was calculated for each observing season ; the results are given in table [ yeartable ] and plotted in figure [ fig3 ] .
the table includes the on - source observing time ( all the time in which the telescope was pointing at mrk421 ) and the statistical significance of the detection for each season . for the hypothesis that the annual rate is a constant , equal to the overall mean rate for the complete 14-year data set ,
the @xmath19 value is 1022 for 13 degrees of freedom , corresponding to a chance probability of @xmath20 . during
some observing seasons mrk421 was the brightest tev source in the sky , whereas in others it was barely detectable with the whipple telescope .
[ cols="^,^,^,^ " , ] [ yeartable ] figure [ fig4 ] ( bottom ) shows the gamma - ray rate averaged over each darkrun ( monthly ) for the 14 years of data . with this unique dataset , it is possible to look for long - term temporal changes in flux levels , slow periodicities and correlation with experiments at other wavelengths with long time constants .
the discrete autocorrelation function ( dacf ) allows the study of the level of autocorrelation in unevenly sampled datasets @xcite without any interpolation or addition of artificial data points .
it can reveal the presence of periodicity ( intrinsic or spurious ) in the time series .
the dacf for the gamma - ray dataset for mrk421 is plotted in figure [ fig5 ] ; it shows no evidence for periodicity at any period between 2 months and 7 years .
we correlated these whipple 10 m observations with archival data taken at other wavelengths for part or all of the epoch 19952009 .
the most complete overlap was with the rossi x - ray timing explorer ( rxte ) database for the entire 14 years @xcite .
there was also significant overlap with milagro @xcite .
there was partial overlap with observations by fermi - lat @xcite and with data from three radio agn monitoring programs ( metshovi radio observatory @xcite , the university of michigan radio observatory @xcite , and mojave @xcite ) .
the correlation with the x - ray data from rxte will be considered here ; the correlation with milagro , fermi - lat and the radio programs will be the subject of a separate publication .
rxte has three non - imaging x - ray detectors , one of which is the all - sky monitor ( asm ) ; it was launched in december 1995 and was operational for the period of the whipple 10 m monitoring program described here .
the asm scanned nearly 80% of the sky in each orbit and was sensitive to x - rays of energy 210 kev .
systematic uncertainties in the asm rate were large , but relative flux estimates were available for month - scale monitoring . in the asm database @xcite , each data point represents the one - day average of the fitted source fluxes from a number ( typically 510 ) of individual asm dwells ( 90 seconds each ) .
the selection criteria for the observations are described in @xcite .
the data are quoted as nominal 210 kev rates in asm counts per second , where the crab nebula rate is about 75 counts per second .
the daily rates were combined ( weighted average ) in @xmath321-day intervals to reduce statistical errors .
the gamma - ray rates for the 19952009 mrk421 whipple observations were compared with the x - ray flux from the asm .
the light curves for the x - ray and gamma - ray data on mrk421 from december 1995 to may 2009 are plotted in figure [ fig4 ] ( top and bottom , respectively ) .
the data are plotted for the whipple 10 m darkrun periods ( @xmath21 for each season ) in which both x - ray and gamma - ray data were taken , i.e. , the x - ray data were averaged for each period in which whipple gamma - ray data were available ( usually 21-day periods ) .
the rxte asm data ( 19952009 ) show evidence for emission correlated with the gamma - ray data on monthly and yearly timescales , as shown in figure [ fig6 ] . in previous papers @xcite , the correlation on short timescales
was treated in detail . here , we concentrate on the longer timescales which are uniquely available in our large database .
the correlation coefficients are @xmath22 and @xmath23 for the monthly- and yearly - binned data , respectively .
there is general agreement between these values and those obtained in previous studies of parts of this database @xcite .
the strong correlation suggests that the x - rays and gamma rays are emitted by the same population of electrons in the same region of the jet . during one flaring episode , a @xmath24 ks time lag of the tev flare with respect to the 24 kev x - ray band
was observed @xcite . for the 14-year data set , the discrete correlation function ( dcf ) between the gamma - ray and x - ray monthly - averaged fluxes for time lags ( gamma ray relative to x - ray ) between @xmath25 and @xmath26 days is shown in figure [ fig7 ] .
the tev and kev fluxes are seen to be significantly correlated ( with a dcf value of 0.82 ) for a time shift shorter than 30 days ; the maximum of the dcf is centered at a lag of ( @xmath27 ) days .
mrk421 is the brightest agn at tev energies ( on average ) and was the first such agn to be discovered at tev energies .
mrk421 is thus the ideal candidate for the detailed study of the variability of this type of object , with the caveat that it may have a greater degree of variability than the norm for such objects on all timescales .
it is remarkable also in being the closest tev - emitting agn .
it is certainly the most studied of the known tev blazars .
most of the variations seen in mrk421 on the various timescales have been seen also in other agn , but the statistics are sparse .
it is probably the youngest such agn .
because it is so close , it has the lowest spectral distortion due to absorption by pair production with the extragalactic background light .
it is difficult to compile a large data sample from a variety of observatories , because observing modes may be quite different and the relative sensitivity of the experiments is often not well - defined .
the whipple 10 m database is comparatively uniform and covers a long time interval ; during the period 19952009 , the system was operated in a relatively stable configuration of telescope and camera @xcite .
it is thus useful to use this database to investigate the statistical distribution of the time variability of mrk421 and , by inference , that of other agn .
if mrk421 is typical of all ( or a subclass of ) tev agn , then we can use these distributions to predict how often a more distant agn might flare or be above a certain level of brightness .
four different timescales are considered : yearly , monthly , daily and run - by - run ( half - hourly ) .
yearly variations : as seen in figure [ fig8 ] , there are significant variations on a yearly timescale , with mean annual rates ranging from 0.23 crab units to 1.86 crab units .
mrk421 was brighter than any other tev source in the sky during most of the 20072008 observing season ( see figure [ fig2 ] ) ; this was also the case for the 20002001 season .
monthly variations ( figure [ fig9 ] ) : there are clearly large variations on this timescale over the full 14 years of data .
there were 19 months ( out of a total of 95 months ) in which the average rate was at least 1 crab unit .
nightly variations : the distribution of the average rate for each of the 783 nights of observation is shown in figure [ fig10 ] . the duration of the observations on which the average is based varied from night to night as observing priorities , weather , etc . , changed .
there were 49 nights ( out of a total of 783 nights , 6.3% ) for which the average rate was at least 2 crab units .
run variations : observing runs were generally of 28-minute duration .
the distribution of the run - by - run rates is shown in figure [ fig11 ] .
this distribution is not a completely unbiased sample , since for a subset of the data there was a tendency for the observers to continue observations when the initial observed signal was high .
there is therefore some bias towards including observations when mrk421 was in a high state ( there is no bias against runs where the signal was low , as all observations were included in the analysis irrespective of the apparent flux level ) .
however , when only the first run in each night is considered , this bias was not found to be strong , as can be seen in figure [ fig12 ] which shows the flux using all the data satisfying weather , hardware and elevation criteria and a smaller set of data where the fluxes are based on only the first run of each night . in later years , this bias was largely removed as observations were concentrated on just mrk421 .
34% of the runs , taken at random , show mrk421 to be at an emission level of @xmath28 crab unit , and 3.5% of the runs have levels above 3 crab units .
the highest fluxes ( @xmath29 crab units ) were observed in sub - run intervals of 25 minutes ( see next section ) .
if mrk421 is typical ( which is probably not the case ) , these distributions could be used to predict the expected level of variation in other tev - emitting agn , scaling by the ratio of their average tev gamma - ray intensity to that of mrk421 ( e.g. , for a tev - emitting agn one - third as bright as mrk421 on average , @xmath33.5% of the runs taken on it might be expected to exhibit flaring above the 1 crab level ) .
mrk421 is characterized by significant flux variations on timescales of a few minutes to years and perhaps longer . during the 14 years of monitoring
reported here , there were three instances in which the variations were sufficiently strong that they were regarded as `` flares '' and were so reported in subsequent publications @xcite .
although the timescales were quite different , the common factor was that in each case the observed flux reached a flux level in excess of 10 crab in a bin of @xmath30 minutes . for completeness
, the salient features of these three instances of exceptional variability are summarized here .
the particular value of these observations was that the signal - to - noise ratio ( number of gamma - ray events to background events ) was sufficiently large that the spectrum of gamma rays could be determined . in the normal operation of observing a relatively weak source with a single cherenkov telescope ,
a parallel series of observations of an off - source region must be taken before a reliable spectrum can be derived .
hence , in this long - term monitoring program , where the emphasis was on determining the lightcurve and off - source measurements were not routinely made , the spectrum could not be reliably derived .
\(i ) during the 19951996 season , mrk421 was relatively quiet ; the mean gamma - ray rate for the season was @xmath31 crab , with a significance of @xmath32 for a total exposure of 51.8 hours ( figure [ fig4 ] ) .
there were two remarkable outbursts of tev gamma rays just eight days apart @xcite .
the first of these occurred on 7th may 1996 ( mjd 50210 ) , when over the course of 2.5 hours the flux increased by a factor of five and reached a maximum rate of @xmath310 crab ( @xmath330 times the average flux for that season ) .
the observations had to be terminated while the flux was still rising due to increasing moonlight so that the peak of the flare was probably not observed . however , on the following night the flux was down to @xmath33 crab , just above the season average . a second flare of lower intensity , but unusual for its very short duration , occurred on 15th may 1996 ( mjd 50218 ) , when a rise and fall within @xmath330 minutes was observed .
the spectrum of the large flare was reported in @xcite as @xmath34 s@xmath35 @xmath36 tev@xmath35 , where the first error is statistical and the second systematic .
this spectrum was confirmed in @xcite , where slight evidence for curvature was found .
\ii ) in 20072008 , the whipple monitoring program again found mrk421 in a strong flaring state ; on 2nd may 2008 ( mjd 54588 ) , the flux exceeded 10 crab @xcite in a 5-minute interval ( figure [ fig13 ] ) .
it appeared that the observation was during the declining phase of a flare whose maximum was not observed .
observations with the veritas telescope on the following night confirmed the high level of gamma - ray activity @xcite
. on 3rd may ( mjd 54589 ) , the whipple telescope and veritas observed mrk421 @xcite , and the light curves are shown in figure [ fig14 ] , binned in 2-minute intervals .
this is the only instance in which observations of the same target with independent imaging atmospheric cherenkov systems were carried out simultaneously with sufficiently high signal - to - noise ratio to permit a meaningful comparison of the gamma - ray rates .
the dcf between the whipple ( single 10 m telescope with energy threshold @xmath3 400 gev ) and veritas ( the array of four 12 m telescopes with energy threshold @xmath3300 gev ) rates , shown in figure [ fig14]c , is an interesting confirmation of the validity of the atmospheric cherenkov technique .
however , because the observations were at large zenith angles , it was difficult to derive a reliable spectrum from the whipple telescope observations .
for the `` very high state '' , the spectrum measured with veritas was fitted using a power law with exponential cutoff model @xcite , of the form @xmath37 yielding normalization constant @xmath38 s@xmath35 m@xmath39 tev@xmath35 at energy @xmath40 tev , spectral index @xmath41 , and cutoff energy @xmath42 tev @xcite . unlike previous observations
, this tev activity did not have a counterpart at x - ray energies and was thus an `` orphan '' flare .
\(iii ) the most notable variability in the mrk421 flux level was observed in 2001 , with an exceptionally bright long - term gamma - ray high state lasting from january to may ( mjd 5192852053 ) . during the months of january and march ,
the flux was highly variable and was four to five times the average flux ; in february , it increased even further ( see figure [ fig15 ] ) .
it is probably wrong to characterize this activity as a flare since it was more like a series of flares .
the source reached a maximum peak flux of @xmath313 crab on 27th february ( mjd 51967 ) .
on one night during this high tev flux state , the detection significance was @xmath43 for a four - hour exposure with an average flux during the four hours of @xmath44 crab . for the 20002001 observing season
, the energy spectrum of mrk421 in the 0.388.2 tev energy range was measured on timescales of a month over a large range of flux states @xcite and fitted using a power law with exponential cutoff model @xcite .
the position of the cutoff energy @xmath45 is highly correlated with the photon index @xmath46 . to test for spectral variability ,
the cutoff energy was fixed at @xmath47 tev . between november 2000 and april 2001 ,
the tev gamma - ray energy spectrum of mrk421 hardened from a photon index @xmath48 at the onset of the large flare to @xmath49 at the peak flux state in late february 2001 @xcite and then softened during the decay of the flare event .
this month - scale flare provides evidence for a shift in the spectral energy distribution to higher energies in the gev - tev band during the peak of the flaring episode . for shorter flaring events , however , there is not always a clear correlation between the tev gamma - ray rate and the spectral shape .
a detailed study was made of one week of data in march 2001 @xcite during which there were several short tev flares and there was significant simultaneous coverage in the x - ray and optical bands . during the flare on 19th march ( mjd 51987 )
, the tev signal was found to lag the x - ray signal in the 14 kev band by @xmath50 minutes .
however , the lag was less at higher x - ray energies .
no correlation was found between the tev and optical signals .
atmospheric cherenkov telescopes are ideally suited for studying agn , but observations are most sensitive on clear dark nights .
the large collection areas associated with the technique make them particularly suited to the study of time variations .
because of their small fields of view , they are generally limited to the study of just one agn at a time . the profusion of sources and the small number of sensitive telescopes available makes long - term monitoring of a particular source difficult .
the whipple 10 m telescope , while lacking the flux sensitivity of the major new observatories , h.e.s.s . ,
magic and veritas , was devoted over a 14-year timespan to observations of mrk421 ( along with a small number of other agn ) and thus provided a unique database .
the whipple 10 m gamma - ray telescope has since been mothballed in the summer of 2011 , after 43 years of operation .
mrk421 is the brightest and most studied blazar that is an emitter at tev energies .
the wealth of observational data makes it an ideal target for modeling the possible emission mechanisms .
however , it is not yet clear to what extent it can be regarded as the archetypal tev - emitting agn . given that a large number of blazars are only observed when they flare , it is useful to know the duty cycle of the flaring activity in order to estimate the chances of detection in any planned observing campaign .
although emission at x - ray and tev gamma - ray wavelengths has been shown to be correlated , emission at other wavelengths ( radio , optical , infrared ) , which can be studied with ground - based telescopes , is not strongly correlated .
hence it is difficult to plan a multiwavelength campaign to observe flaring activity that could be triggered by a non - gamma - ray signal .
only a small number of blazars have been detected by the whipple 10 m telescope ( mrk421 , markarian 501 ( mrk501 ) , 1es2344 + 514 , h1426 + 428 and 1es1959 + 650 ) .
apart from mrk421 and mrk501 , the average tev flux level measured for these blazars is extremely low .
for example , the average rate observed by the whipple 10 m telescope from h1426 + 428 during 2000 and 2001 was only ( @xmath51)% of the crab nebula rate @xcite . 1es2344 + 514 and 1es1959 + 650 were only detected by the 10 m telescope because of strong flaring activity @xcite .
it is therefore difficult to confirm the predictions of flaring frequency using 10 m data alone . as more tev blazars
are identified using data from h.e.s.s . , magic , and veritas , and their flaring statistics determined in long exposures , it will be possible to ascertain to what extent mrk421 is typical of tev - emitting agn .
this research is supported by grants from the u.s .
department of energy office of science , the u.s . national science foundation and the smithsonian institution , by nserc in canada , by science foundation ireland ( sfi 10/rfp / ast2748 ) and by stfc in the u.k .
we acknowledge the excellent work of the technical support staff at the fred lawrence whipple observatory and at the collaborating institutions in the construction and operation of the instruments .
ulrich , m.h .
et al . 1997 , ann . rev .
astron . astrophys . , 35 , 445 | the variability of the blazar markarian 421 in tev gamma rays over a 14-year time period has been explored with the whipple 10 m telescope .
it is shown that the dynamic range of its flux variations is large and similar to that in x - rays . a correlation between the x - ray and tev energy bands
is observed during some bright flares and when the complete data sets are binned on long timescales .
the main database consists of 878.4 hours of observation with the whipple telescope , spread over 783 nights .
the peak energy response of the telescope was 400 gev with 20% uncertainty .
this is the largest database of any tev - emitting active galactic nucleus ( agn ) and hence was used to explore the variability profile of markarian 421 .
the time - averaged flux from markarian 421 over this period was @xmath0 crab flux units .
the flux exceeded 10 crab flux units on three separate occasions .
for the 20002001 season the average flux reached 1.86 crab units , while in the 19961997 season the average flux was only 0.23 crab units .
agn , tev gamma rays , markarian 421 |
in the study of the phase behavior of colloidal mixtures there are two very distinct approaches . from a theoretical point of view
one would like to keep the description of the mixture as simple as possible in order to be able to calculate the phase diagram using well established theoretical tools .
furthermore , it is possible to _ understand _ the main physics , i.e. the phase behavior , bulk correlation functions , or inhomogeneous density distributions , in simple terms .
one example of such an theoretical approach is the asakura - oosawa - vrij ( aov ) model for colloid - polymer mixtures.@xcite within the aov model , colloids are treated as hard spheres , while polymer behave like hard spheres only when interacting with colloids .
the polymer - polymer interaction vanishes .
based on this model , one finds that the phase diagram of a colloid - polymer mixture , assuming perfectly monodisperse colloids and polymer , resembles that of a simple fluid , and the reservoir polymer density plays the role of an inverse temperature .
one , relatively simple but successful , theoretical approach for treating the phase behavior of this model colloid - polymer mixture is the free - volume theory ( fvt ) that assumes a low concentration of the depletion agent , the polymer.@xcite a very different picture of the phase behavior of colloidal mixture originates from experimental studies . while in some cases it is possible to prepare model systems with almost monodisperse colloids and polymer,@xcite the generic case of a colloidal sample is polydisperse due to imperfections of the synthesis process .
polydispersity can be found both in size and in shape distributions . from the theoretical point of view
, polydispersity is often considered as a defect , because it complicates calculations . however , polydispersity can give rise to interesting behavior , which can be utilized in experiments,@xcite if the role of polydispersity is sufficiently well understood .
one example is the study of the glass transition , where polydispersity in model hard colloidal spheres has the role as to suppress the freezing transition at high density of colloids so that a glass can form.@xcite there are various very different ways to achieve a successful description . for hard and permeable spheres salacuse and
stell@xcite showed a thermodynamic approach based on the percus - yevick equation of state and the more accurate approach by mansoori _
et al_. the polydispersity in the chain - length of self - excluding polymer has been treated in ref . .
polydispersity has been incorporated in the theory in different manner .
near - monodisperse distributions have been treated in a perturbative fashion@xcite and bimodal size distributions has been studied in ref . .
in the present study we wish to increase the understanding of effects that can be caused by polydispersity further . to this end
we consider a mixture of monodisperse colloidal spheres with a polydisperse depletion agent .
we keep the theory simple by assuming that the depletion agent is efficient and can produce a sufficiently strong depletion attraction between colloidal spheres at rather low concentrations . in this limit we can generalize the free - volume theory within the framework of fundamental measure theory@xcite ( fmt ) to cases of polydisperse size or shape distributions . as an application of our general and powerful approach we study the phase behavior of a ternary mixture of ( monodisperse ) colloidal spheres with two polydisperse polymer components with distinct size difference .
our study shows that the distribution of the depletion agent can be employed to tune the phase behavior of colloidal mixtures in a wide range and it is possible to generate a third fluid phase . beside the low density ( in colloids ) gas and a high density liquid , we find a low density liquid phase .
this is similar to findings in systems in which two length scales are introduced in the inter - particle interaction potentials.@xcite the paper is structured as follows . in section [ sec::theory ]
we recall the theoretical basics of free - volume theory within the framework of fmt .
we extend the approach of ref . to treat discrete and continuous distributions of the depletion agent . in section [ sec::distributions ]
we discuss a few distribution functions which are frequently used in studies of polydisperse systems and show that for infinitely thin needles or platelets various types of size distributions lead to explicit expressions for the effective free - volume fraction @xmath0 . in section [ sec::results ]
we show and discuss as an application of our theory the phase diagrams for the case of a ternary mixture of ( monodisperse ) colloidal spheres and two polydisperse polymer components .
we discuss the results and conclude in section [ sec::discussion ] .
we study mixtures of colloidal spheres immersed in a solvent of depletion agent .
the latter component can possess a polydisperse size or shape distribution . here
we are mainly interested in mixtures where the depletion potential between the spheres is strongly attractive already at rather low concentrations of the depletion agent . in this case
we can employ the description of asakura , oosawa@xcite and vrij@xcite for the mutual particle interactions , which keeps the model simple .
the colloid - colloid and the colloid - depletion agent interactions are modeled as hard - core interactions , while the mutual interaction between particles of the depletion agent vanish .
it has been shown @xcite that the phase behavior of the asakura - oosawa - vrij model is closely related to the free - volume theory , in which the excess free energy of the mixture is expanded up to first order in the density of the depletion agent .
this linearization of the excess free energy is equivalent to a vanishing interaction among particles of the depletion agent . as fmt@xcite provides a good framework for the free - volume theory we first recall its main ideas as presented in refs . and for discrete mixtures and generalize it to continuous distributions of the depletion agent . in order to describe the phase behavior of colloidal mixtures one has first to choose the thermodynamic ensemble . for our purpose
a system with a fixed number of big spherical colloidal particles @xmath1 , which is coupled to a reservoir of a depletion agent at chemical potential @xmath2 or fugacity @xmath3 , respectively , represents the best choice.@xcite hence , the system is treated in a semi - grand canonical ensemble .
the helmholtz free energy separates into two terms @xmath4 the first term is the helmholtz free energy of pure colloids . by taylor expansion of the integrand of the second term in eq .
( [ eqn::semigcpotential ] ) up to first order in @xmath3 we obtain an expression for the volume available to particles of the depletion agent in a system of @xmath1 colloids at low concentrations of the depletion agent , the so - called free volume denoted by @xmath5 . in this limit of small fugacity of the depletion agent ,
we can rewrite eq .
( [ eqn::semigcpotential ] ) into @xmath6 where @xmath7 is the reservoir density of the depletion agent . within the aov model
the free - volume fraction @xmath8 depends only on the geometry of the depletion agent , e.g. the size ratio @xmath9 in the case of spherical polymer , and on the packing fraction of the colloidal spheres @xmath10 . for the calculation of the phase diagram of the colloidal mixture we require the chemical potential of the colloid component @xmath11 and the osmotic pressure @xmath12 . in the bulk
two phases i and ii coexist if they are in chemical and mechanical equilibrium , i.e. if @xmath13 and @xmath14 , respectively .
the chemical potential @xmath11 and the pressure @xmath12 follow from the helmholtz free energy via@xcite @xmath15 and @xmath16 the chemical potential and the pressure of the pure reference system , labeled @xmath17 and @xmath18 , are described accurately by the carnahan - starling expressions in the fluid phase . to assure thermodynamic consistency the free - volume fraction @xmath8
must also be calculated from the equation of state of the reference system , i.e. the carnahan - starling equation of state.@xcite besides coexisting fluid states , i.e. points on the binodal line in phase diagram , we calculate the spinodal line which separates metastable states from thermodynamically unstable ones . for thermodynamically unstable states the second derivative of the helmholtz free energy w.r.t . the colloid packing fraction @xmath19 is negative and for points on the spinodal line it vanishes ,
i.e. @xmath20 . for the solid - fluid coexistence we need the chemical potential and equation of state of a reference ( fcc ) hard - sphere crystal , which is described by @xcite @xmath21 with @xmath22 , the packing fraction for close packing , and @xcite @xmath23 for discrete mixtures of two species the excess helmholtz free energy density @xmath24 is constructed from two sets of weighted densities @xmath25 and @xmath26 , one for the colloids @xmath27 and one for the depletion agent @xmath28.@xcite in homogeneous bulk fluids @xmath29 labels four scalar - weighted densities , which can be identified as the scaled - particle theory variables @xmath30 of the @xmath31-component mixture@xcite @xmath32 with the geometrical measures of component @xmath33 : volume @xmath34 , surface area @xmath35 , integrated mean curvature @xmath36 and euler characteristics @xmath37 .
an explicit expression for the free energy density of the aov mixture in the limit of low density of the depletion agent can be obtained by expanding the hard - sphere ( hs ) mixture free energy with weighted densities @xmath38 up to linear order in the density @xmath39:@xcite @xmath40 in order to relate the terms in eq .
( [ eqn::phiao ] ) to the free - volume fraction @xmath8 we transform the excess helmholtz free energy density , which in eq .
( [ eqn::semigcpotential ] ) is given in the semi - grand canonical ensemble , into the canonical ensemble and obtain in terms of the colloid packing fraction @xmath19 and the density @xmath39 @xmath41 by direct comparison of eq .
( [ eqn::phiao ] ) and eq .
( [ eqn::phisemigrandcan ] ) we obtain a fmt expression for the free - volume fraction@xcite @xmath42 the partial derivatives of the free energy w.r.t .
the weighted densities are thermodynamic quantities of the ( pure ) colloids , namely , the pressure @xmath18 , the surface tension @xmath43 at a planar hard wall and the bending rigidities @xmath44 and @xmath45 , which describe the effect on the free energy due to curved surfaces .
@xcite the depletion agent enters the expression for the free - volume fraction @xmath8 only through its geometrical measures @xmath46 . for the free energy density we employ rosenfeld s original formulation of fmt@xcite and the more accurate white bear version from refs .
@xcite @xmath47 which is based on the boublik - mansoori - carnahan - starling - leland ( bmcsl ) equation of state@xcite @xmath48 it would also be possible to base the free - volume theory on a recent , slightly more consistent , mixture equation of state.@xcite fmt is a mixture theory by construction . for mixtures of colloids and ideal depletion agent we can employ parameters @xmath49 to parameterize the geometry of particles of the depletion agent components .
although it is straightforward within fmt to treat general mixtures of @xmath50 colloid and @xmath51 depletion agent components , we will in the following restrict our studies to the simpler case of a single colloid species with packing fraction @xmath19 .
for the depletion agent we start by considering @xmath51 components . in the semi - grand canonical ensemble
the excess helmholtz free energy density is , analogous to eq .
( [ eqn::phisemigrandcan ] ) , given by @xmath52 and the set @xmath53 of geometrical measures is characterized by the set of parameters @xmath54 . to study the influence of polydispersity in the depletion agent on the phase behavior
it might be sufficient to consider a discrete mixture , however we prefer to introduce continuous distributions , which can represent experimental systems more accurately than a discrete distribution . in the following we restrict the set @xmath53 to depend on a single parameter @xmath55 only .
we introduce @xmath56 as a continuous distribution of the depletion agent in the reservoir , and require that @xmath57 .
the density distribution of the depletion agent follows directly : @xmath58 . by specifying the fundamental geometrical measures of the depletion agent as function of @xmath55 we obtain , similar to ref .
, @xmath59 ) = \phi(\eta_c ) \nonumber\\&&- \rho_d^r \int { \textrm{d}}q \ , { d}(q ) \alpha(\eta_c , v_d(q ) , a_d(q ) , c_d(q ) , x_d(q ) ) , \label{eqn::phipolydisperse}\end{aligned}\ ] ] which is a functional of the distribution @xmath60 . note that the precise meaning of @xmath55 is unspecified so far and can refer to a size ratio , in the case of spherical mixtures , or to a parameter that specifies the shape of the depletion agent in a more complicated manner .
we return to this point later .
analogous to eq .
( [ eqn::freeenergysemigc ] ) we call the integral on the r.h.s . of eq .
( [ eqn::phipolydisperse ] ) the effective free - volume fraction @xmath0 and rewrite eq .
( [ eqn::phipolydisperse ] ) as @xmath61 ) = \phi(\eta_c ) - \rho_d^r \alpha_{\textrm{eff}}(\eta_c ; [ { d } ] ) .
\label{eqn::phipolydispersealphaeff}\ ] ]
to study the phase behavior of a mixture of spherical colloids and a polydisperse depletion agent we focus on three frequently used distributions , namely the schulz ( @xmath62 ) , gaussian ( @xmath63 ) and the hat - like ( @xmath64 ) distributions .
these distributions are characterized by two parameters corresponding to its first and second moment .
the average asymmetry ratio is given by @xmath65 , with the colloid diameter @xmath66 and a length - scale of the depletion agent @xmath67 .
the parameter @xmath68 describes the degree of polydispersity : the limiting case @xmath69 equals the monodisperse fluid and by decreasing the value of @xmath68 the distribution is broadened .
the distributions under consideration here lead to explicit expressions for the effective free - volume fraction @xmath70 , if the geometries of the depletion agent is simple . for more complicated shapes we have to perform the integration numerically . in terms of @xmath71 and
@xmath68 the distributions considered here are given by @xmath72 , \label{eqn::gaussiandist}\\ { d}_{h}(q;\bar q , z ) & = & \frac{z}{2 } \theta\left(q-\bar q+z^{-1}\right)\nonumber\\ & & \times\theta\left(-q+\bar q+z^{-1}\right ) , \quad \quad \bar q z > 1 .
\label{eqn::hatlikedist}\end{aligned}\ ] ] it is evident that @xmath73 is required in all distributions and the integration in @xmath74 is performed in the limits from 0 to @xmath75 . in the case of a gaussian distribution one
can obtain simpler explicit expressions by imposing the full integration range @xmath76 to @xmath75 .
the result , @xmath77 , is similar to that of the cut gaussian distribution @xmath78 @xmath79 obviously , it is possible to consider different size distributions such as the log - normal distribution , which decays slower for large values of @xmath55 than the distributions considered here . the equations for the chemical potential @xmath11 and the osmotic pressure @xmath12 from eqs .
( [ eqn::chemicalpotential ] ) and ( [ eqn::osmoticpressure ] ) for polydisperse size distributions contain integrals over the distribution .
for the calculation of the phase diagram these equations must be solved simultaneously . in general this must be done numerically . in the case of one- and two - dimensional depletion agents , such as infinitely thin rods or platelets , their geometry is sufficiently simple so that the effective free - volume fraction @xmath0 can be calculated explicitly .
infinitely thin needles represent an one - dimensional depletion agent .
infinitely thin platelets represent a two - dimensional depletion agent . here
we consider both disk- and hexagon - shaped platelets .
the geometric measures for these geometries are given in tab .
[ tab::geometricmeasures ] . in appendix [ app::explicitexpressions ]
we show explicit expressions for @xmath0 for one- and two - dimensional depletion agents and the distributions given above . as aforementioned ,
we obtain the thermodynamic quantities of the pure hard - sphere fluid that enter the free - volume fraction @xmath0 from the excess free energy density @xmath80 of the white bear version of fmt.@xcite these quantities are the pressure @xmath81 , the surface tension @xmath82 and bending rigidities @xmath83 and @xmath84.@xcite in the present study we restrict the product @xmath85 to sufficiently large values . for the hat - like distribution we require @xmath86 which ensures that @xmath87 . in the case of the gaussian distribution @xmath77 we require @xmath86 because for small values of @xmath85 the contribution of the distribution from values @xmath88 increases , e.g. for @xmath89 and @xmath90 we obtain @xmath91 . in the case of two - dimensional depletion agents
, the geometric measures can differ for different geometry of the depletion agent : for disk- and hexagonal platelets @xmath92 and @xmath93 differ see tab .
[ tab::geometricmeasures ] . for that reason
we show the result for @xmath0 in appendix [ app::explicitexpressions ] in its general form .
.geometrical measures for one- and two - dimensional depletion agents .
all lengths are measured in units of the diameter @xmath66 of the spherical colloids .
we consider needles of length @xmath94 and radius @xmath95 and platelets with radius @xmath96 and thickness @xmath97 .
the value of @xmath55 plays the role of a size ratio and compares the size of the depletion agent to that of a colloid . [
cols="^,<,<,^,^,^,^",options="header " , ]
while our approach developed in section [ sec::theory ] is general and allows one to study various systems , we want to present only one application .
we study the phase behavior of a ternary mixture of ( monodisperse ) colloidal spheres with two polydisperse polymer components of distinct sizes .
both polymer components possess a size distribution which we describe by schulz distributions , which in total lead to @xmath98 with a mixing parameter @xmath99 and two averaged size ratios @xmath71 and @xmath100 . for simplicity
we assume the width of both parts of the distribution to be equal .
the polymer reservoir packing fraction is given by @xmath101 , where @xmath102 .
we expect the phase behavior to be most interesting if the average size ratios @xmath71 and @xmath100 are distinct , so that the phase diagrams in the limiting cases @xmath103 and @xmath104 are sufficiently different . to this end
we choose @xmath105 and @xmath106 . in the case of @xmath103 when we obtain a binary mixture with an averaged size ratio @xmath106 , the phase diagram exhibits a stable fluid - fluid phase separation into a colloid poor gas and a colloid rich liquid phase . for the mixing parameter @xmath104 ,
when the averaged size ratio @xmath107 , the fluid - fluid phase separation , which is metastable w.r.t .
to crystallization in the monodisperse case,@xcite can be stabilized by a sufficient degree of polydispersity.@xcite before we study the full phase diagram of the ternary mixture , we determine which value of @xmath108 is required in order to stabilize the fluid - fluid phase separation for @xmath104 . in this section
we study the phase diagram of a binary colloid - polymer mixture , which corresponds to the limiting case @xmath104 , with an averaged size ratio of @xmath107 .
in contrast to the aforementioned examples of one- and two - dimensional depletion agents , @xmath0 can not be obtained explicitly .
therefore , we have to determine the phase coexistence by solving two equations , eqs .
( [ eqn::chemicalpotential ] ) and ( [ eqn::osmoticpressure ] ) , numerically . the equation of state of the pure reference system and the free - volume fraction @xmath0 respectively , and hence the chemical potential @xmath11 and the osmotic pressure @xmath12 of the colloids , are based on the free energy density @xmath80 of the white bear version of fmt , eq .
( [ eqn::phiwhitebear ] ) . in fig .
[ fig::phasediagrampolyao](a ) we show the phase diagrams of polydisperse aov mixtures as we vary the degree of polydispersity . for the size ratio @xmath107 and a narrow distribution of the polymer ,
the fluid - fluid phase separation remains metastable similar to the monodisperse case .
we show the phase diagram for @xmath109 ( full lines ) in fig .
[ fig::phasediagrampolyao](a ) and the corresponding size distribution in fig .
[ fig::phasediagrampolyao](b ) . upon decreasing @xmath68 , which broadens the size distribution ,
the fluid - fluid and the fluid - solid coexistence lines come closer to each other and almost touch for @xmath110 ( not shown ) . by further increasing the degree of polydispersity
the fluid - fluid coexistence becomes stable , as we show in fig .
[ fig::phasediagrampolyao](a ) for @xmath111 ( dashed - dotted lines ) and @xmath112 ( dotted lines ) .
the corresponding size distributions are shown in fig .
[ fig::phasediagrampolyao](b ) .
the critical point is shifted to lower colloid packing fractions @xmath19 , as the value of @xmath68 decreases .
the polymer distribution in the system differs from that in the reservoir , @xcite which is a phenomenon known as fractionation . for spherical polymer
we find results equivalent to those of ref . .
at low packing fractions of the colloids the size distribution of ( spherical ) polymer follows closely the reservoir distribution . as the colloid packing fraction increases
, the maximum of the polymer size distribution decreases significantly and moves towards smaller polymer radii . in principle
we observe a qualitatively similar but less pronounced behavior also in the case of platelet - like or rod - like depletion agents .
the maximum in the size distribution of the depletion agent becomes smaller and moves to smaller sizes as the colloid packing fraction in the system increases .
this is to be expected , because one- or two - dimensional depletion agents require less free volume to fit between colloids than spherical polymer .
we have confirmed that a treatment of the phase behavior completely based on the py compressibility equation of state yield qualitatively similar results . in the present approach
this is achieved by employing the free energy density @xmath80 of rosenfeld s original formulation of fmt to evaluate @xmath0 and the pressure and chemical potential for the pure reference system .
the results in refs . , , and are somewhat in between these two predictions as their approach employs a hard - sphere reference system based on the carnahan - starling equation of state and a free - volume fraction based on the py compressibility equation of state , which for spheres is equivalent to the scaled - particle theory expression . in this section
we briefly address the question whether polydispersity , which describes a shape distribution , can also stabilize the fluid - fluid phase separation .
one can think of the shape polydispersity in two different ways .
on the one hand it can be thought as describing effects due to imperfections in the synthesis of colloidal particles . on the other hand , and this is more relevant in the present context , it can be seen as a simple model for the effect that polymer coils are not rigid and can change their average shape when squeezed in between colloids . to this end
we consider ellipsoids , with half - axes @xmath113 and @xmath114 , for which the geometrical measures can be obtained explicitly.@xcite this case is relatively simple as a single parameter @xmath55 characterizes the shape while @xmath115 is fixed .
obviously , more complicated scenarios are possible
. note that for this choice of half - axes the volume @xmath116 is kept fixed , while the surface area @xmath117 and integrated mean curvature @xmath118 depend on @xmath55 .
the euler characteristic @xmath119 , independent of @xmath55 . here
, the parameter @xmath55 describes the degree of deviation of the shape from a sphere : for small values of @xmath55 we obtain lens - shaped particles ( oblates ) , while for large values of @xmath55 the particles become cigar - shaped ( prolates ) .
similarly to the case of size polydispersity , we observe that an increasing degree of shape polydispersity stabilizes the fluid - fluid phase separation w.r.t . fluid - solid decomposition .
[ fig::shape ] we show an example for a narrow shape distribution with @xmath109 ( full line ) , in which case the fluid - fluid critical point is metastable , and one for a broad distribution with @xmath112 ( dotted line ) . for @xmath112 we find a stable fluid - fluid phase separation .
it seems that polydispersity in general favors a stable fluid - fluid phase separation because the polydisperse depletion agent can fill the free volume of the system more effectively than a monodisperse one . even if the bigger particles of the distribution can not find free volume in the system
, the smaller ones still do . in terms of the depletion potential
this would result in a longer range of the effective attraction between colloids,@xcite giving rise to a more negative effective second virial coefficient @xmath120 which increases the tendency for fluid - fluid phase separation.@xcite this effect seems to be robust against details of introducing polydispersity and can also be observed for a depletion agent with @xmath121 and @xmath114 , where all geometrical measures , except of the euler characteristics , depend on @xmath55 . based on our findings in section [ sec::sizepoly ]
we choose @xmath122 for the bimodal polydisperse polymer distribution . before we show the calculated phase diagrams , it is interesting to compare the two _ stable _ critical points corresponding to the limiting cases , i.e. @xmath104 and @xmath103 .
the critical points can be found at very distinct regions of the phase diagram . in fig .
[ fig::criticalpoints ] we show the paths of the critical points through the phase diagram as a function of the mixing parameter @xmath99 .
when we start at @xmath104 , for which only the smaller polymer are present , we find a critical point at @xmath123 and @xmath124 .
as we decrease the value of @xmath99 , the critical point moves towards smaller values of @xmath19 until this critical point becomes metastable . for @xmath125
the critical point no longer exists .
when we start at @xmath103 , for which only the larger polymer are present , we find a second critical point at @xmath126 and @xmath127 . as the value of @xmath99
is increased we find that this critical point moves towards larger values of @xmath19 until it becomes metastable . for @xmath128 this critical point vanishes .
it is most interesting to note that both critical points found in this system , which correspond to critical points of the limiting cases , can exist at the same time for a range of values of the mixing parameter , @xmath129 $ ] . in order to obtain a full picture and determine where these critical points are stable we calculate the full phase diagrams in this regime . in figs .
[ fig::ternaryphasediagrams](a)-(b ) we show phase diagrams in the @xmath19-@xmath130representation for different values of @xmath99 . for @xmath131
[ see fig .
[ fig::ternaryphasediagrams](a ) ] the phase diagram is similar to the monodisperse case studied above .
upon decreasing @xmath99 , the fraction of polymer with average size ratio @xmath100 increases , and the spinodal displays two minima , one corresponding to the stable critical point ( triangle ) and the second minimum corresponding to an additional metastable fluid - fluid phase separation ( diamond ) see fig .
[ fig::ternaryphasediagrams](b ) . further decreasing the value of the mixing parameter to @xmath132 [ fig .
[ fig::ternaryphasediagrams](c ) ] yield a phase diagram with two _ stable _ critical points and a second triple point . at one triple point ,
a gas , a liquid , and a solid phase coexist , while at the second triple point a gas phase and two liquid phases with different densities coexist .
there is a novel low density liquid phase in the region between the two critical points with colloid packing fractions @xmath133 and @xmath134 .
as we vary the value of @xmath99 the triple points move . according to gibbs phase rule a four - phase coexistence between a gas , the low density liquid , the high density liquid and the solid is also possible . for our system
it would be expected for @xmath99 between the values used in fig .
[ fig::ternaryphasediagrams](b ) and ( c ) . in fig .
[ fig::ternaryphasediagrams](d ) for @xmath135 we also observe two critical points .
the critical point with lower colloid packing fraction ( diamond ) is stable .
the second critical point at higher colloid packing fraction ( triangle ) is now in the metastable regime and vanishes upon decreasing the values of @xmath99 . for these systems
one liquid phase and one triple point are stable . further decreasing the values of @xmath99 below approximately 0.8 lead to a shift of the coexistence region to higher values of @xmath136 .
for @xmath137 the fluid - fluid binodal is in the region of the fluid with monomodal polydisperse depletion agent with @xmath106 .
we observe similar behavior for a mixture of colloids and two monodisperse polymer components , i.e. @xmath138 and @xmath139 .
for the same size asymmetries as above , @xmath140 and @xmath141 , the critical point is metastable when @xmath104 , or equivalently @xmath142 .
we observe one stable and one metastable critical point for low colloid packing fraction upon decreasing the value of @xmath99 . in the case @xmath143 and @xmath144 there is a stable critical point for @xmath104 .
decreasing @xmath99 , similarly to the polydisperse case described above , leads to a second stable critical point .
to summarize , we find that both by means of choosing the degree of polydispersity and the size asymmetry we can treat a model fluid which exhibits one or two ( stable ) critical points and one or two ( stable ) liquid phases .
furthermore , we have verified that it is possible to generate two critical points and hence _ three _ fluid phases by using a depletion agent of different than spherical shape .
we have generalized a novel approach to free - volume theory@xcite based on fmt@xcite to treat a depletion agent of arbitrarily shaped particles with a continuous distribution . here , we introduced one parameter , @xmath55 , to specify a continuously varying distribution of depletion agent . to evaluate the phase diagram within fvt an expression for the free - volume fraction
is required . in the case of polydisperse size or shape distributions for the depletion agent
, one can obtain an _ effective _ free - volume fraction , which in general has to be calculated numerically . in the cases of infinitely thin needles or platelets it is possible to get explicit expressions for the effective free - volume fraction for various distributions which are of interest from an experimental and a theoretical point of view .
we give some expressions for @xmath0 in appendix [ app::explicitexpressions ] . beside size polydispersity
, we have also considered _ shape _ polydispersity . to this end
we have calculated phase diagrams of a mixture of colloidal spheres with penetrable ellipsoids with half - axes @xmath145 and @xmath114 .
our results confirm that polydispersity has the tendency to stabilize the fluid - fluid coexistence region w.r.t .
we have discussed this trend , in agreement with earlier studies,@xcite in section [ sec::sizepoly ] for size polydisperse spherical polymer coils . in section [ sec::shapepoly ]
we have shown that similar behavior can be found for the case of shape polydisperse case . upon increasing the degree of polydispersity the fluid - fluid binodal `` crosses '' the fluid - solid coexistence line and one obtains a stable critical point with associated liquid and gas phases and a triple point .
this effect seems to be robust against various ways to incorporate polydispersity , which can be rationalized by the fact that particles of a polydisperse depletion agent can fill the available free volume in a system better than a monodisperse species .
our main result is the study of mixtures of colloids and two species of polydisperse ( spherical ) polymer coils .
the polymer were described by a bimodal distribution consisting of two polydisperse distributions . for the averaged size ratios of the polymer components chosen here , i.e. @xmath146 and @xmath106 , we find in distinct parts of the phase diagram fluid - fluid phase separations in the two limiting cases @xmath104 , when only the smaller polymer species is present , and @xmath103 , when only the larger polymer species is present . for @xmath103
the corresponding fluid - fluid phase separation is stable independent of the degree of polydispersity . in the second limiting case ,
@xmath104 , the fluid - fluid phase separation has to be stabilized by a sufficient degree of polydispersity .
we choose @xmath122 and find for a range of the mixing parameter @xmath99 two stable liquid phases : beside the gas phase , which is diluted in the density of colloids , we find a low density and a high density liquid phase .
the fluid - fluid coexistence region is w - shaped in this case and the phase diagram exhibits two stable critical points and two triple points . for systems for which the inter - particle interactions contains two length scales , such as short - range repulsion and long - range attraction , hemmer and stell@xcite found that two stable critical points may occur .
more directly related to our situation is a system of star polymers@xcite in which similar features in the phase diagrams are reported , as the effective interaction potential is comparable to those of the model of stell and hemmer . for other model fluids
similar effects were observed.@xcite we point out that our findings are not restricted to a polydisperse depletion agent .
we find similar results for a ternary mixture in which we consider one colloid species and two distinct monodisperse species of depletion agent . while often size or shape polydispersity is considered an undesirable side - effect , originating from imperfections of particle synthesis , it is possible to make use of this property by stabilizing fluid - fluid phase separations .
moreover , the degree of polydispersity allows one to adjust the phase diagram . from this point of view
one might speculate about fine tuning the phase behavior of colloidal mixtures by specifying the polydisperse distribution of the depletion agent .
explicit expressions for the effective free - volume fraction in the case of infinitely thin needles for the schulz , gaussian and hat - like distributions , eqs .
( [ eqn::schulzdist])-([eqn::hatlikedist ] ) , read since only @xmath44 and @xmath148 enter these expressions for a one - dimensional depletion agent , there is no difference between an approach based on the py or the bmcsl equation of state . in terms of the colloid packing fraction @xmath19 , average asymmetry ratio @xmath71 and the polydispersity parameter @xmath68
we obtain @xmath150 \sqrt { { z}^{2}- \chi_2 } } } , \\
\alpha_{\textrm{eff},g}'(\eta_c;\bar q , z ) & = & \exp { \left ( { \frac { { z}^{2}{\bar q}^{2 } \chi_2 + \chi_0 \,{z}^{2}- \chi_0\ , \chi_2+{z}^{2}\bar q \chi_1 + { \chi_1}^{2}/4}{{z}^{2}- \chi_2}}\right ) } { \frac { z}{\sqrt { { z}^{2}- \chi_2 } } } , \\ \alpha_{\textrm{eff},h}(\eta_c;\bar q , z ) & = & \frac{z}{4}\sqrt{\frac{\pi}{-\chi_2 } } { \exp{\left ( \chi_0 - \frac{\chi_1 ^ 2}{4 \chi_2 } \right ) } } \sum_{n=1}^{2 } ( -1)^n { \mathrm{erf}}\left ( { \frac { 2 \chi_2 ( \bar q - ( -1)^{n } ) + \chi_1z}{2 z \sqrt { -\chi_2 } } } \right),\end{aligned}\ ] ] where @xmath151 , @xmath152 and @xmath153 . the geometrical measures @xmath92 , @xmath93 and @xmath154 are given in tab .
[ tab::geometricmeasures ] .
@xmath155 with @xmath156 are parabolic cylinder functions . | we study the phase behavior of mixtures of monodisperse colloidal spheres with a depletion agent which can have arbitrary shape and can possess a polydisperse size or shape distribution . in the low concentration limit , considered here , we can employ the free - volume theory and take the geometry of particles of the depletion agent into account within the framework of fundamental measure theory . we apply our approach to study the phase diagram of a mixture of ( monodisperse ) colloidal spheres and two polydisperse polymer components . by fine tuning the distribution of the polymer it is possible to construct a complex phase diagram which exhibits two stable critical points . |
in supersymmetric models , gauge invariance allows interactions that violate lepton- or baryon - number conservation . to avoid them
, a discrete symmetry called r - parity is often imposed .
the r - parity of a particle is defined as @xmath3 , where @xmath4 is its spin , and @xmath5 and @xmath6 are its lepton and baryon numbers .
@xmath7 is @xmath8 for all standard model particles and @xmath9 for all superparticles .
alternatively , it is an interesting exercise to allow r - parity violating couplings and constrain them by experimental observations or cosmological considerations . with r - parity violation
, the lightest superpartner decays into standard model particles .
we will assume that the lightest superpartner is the neutralino .
antiprotons are a rare component of cosmic rays , most probably produced in spallation reactions in the interstellar medium .
the measured antiproton flux limits any additional antiproton production in the galaxy .
we find a powerful bound on the lifetime of the lightest superpartner from the observed flux of cosmic ray antiprotons .
this allows us to bound the r - parity violating interactions that appear in extensions to the minimal supersymmetric standard model .
we find the relic density of the lightest superpartner ( lsp ) using the standard cosmological model .
we then allow these particles to decay via the hadronic r - parity violating operators .
we employ simple models for the production rate and the propagation of the antiprotons .
we can limit the r - parity violating couplings with observations of the antiproton spectrum in cosmic rays .
our results strictly apply when the lifetime of the neutralino is longer than the age of the universe
. however , there are other cosmological limits based on nucleosynthesis in the early universe , recent structure formation , diffuse photon and neutrino fluxes , and spectral distortions in the cosmic microwave background that cover a large fraction of neutralino lifetimes shorter than the age of the universe , and may well completely close the gap between this result and laboratory constraints .
the supersymmetric model we consider is a minimal extension of the standard model with explicit r - parity breaking , defined by the superpotential @xmath10 @xmath11 are generation indices ; @xmath12 and @xmath13 are the su(2)-doublet lepton and quark superfields ; @xmath14 , @xmath15 and @xmath16 are the singlet superfields ; and @xmath17 and @xmath18 are the doublet higgs superfields .
@xmath19 , @xmath20 and @xmath21 are yukawa matrices , and @xmath22 is a free parameter @xcite .
the last two terms in the superpotential violate lepton- and baryon - number conservation explicitly . for three generations , there are 27 @xmath23 couplings and 9 @xmath24 couplings , since @xmath25 is antisymmetric under exchange of the last two indices . supersymmetry is broken softly by a scalar potential that in general contains many unmeasured parameters . for our purposes ,
a simple parameterization suffices .
we assign a common mass scale @xmath26 to the sleptons and squarks at the electroweak scale , we relate the gaugino mass parameters @xmath27 by means of gut relations , and we keep the trilinear soft terms @xmath28 and @xmath29 only for stop and sbottom
. electroweak symmetry is broken at tree level by assigning vacuum expectation values @xmath30 and @xmath31 to the neutral scalar components of @xmath17 and @xmath18 .
the ratio @xmath32 and the pseudoscalar higgs boson mass @xmath33 remain as free parameters .
the total number of parameters is therefore seven .
further details on the class of models we consider are given by bergstrm and gondolo @xcite .
we use the scans of supersymmetric parameter space in bergstrm , edsj , and gondolo @xcite .
they consist of @xmath34 models that satisfy the experimental constraints reported by the particle data group @xcite , plus the bounds on the higgs sector and chargino masses from lep ( we update to lep2w @xcite ) and the constraints on the @xmath35 rate from cleo @xcite . the scans span values of @xmath36 and @xmath37 up to 5 tev , of @xmath26 up to 3 tev , of @xmath38 and @xmath39 up to @xmath40 , of @xmath33 up to 1 tev , and of @xmath41 between 1 and 50 .
this range extends even beyond the parameter region usually considered as plausible .
we do not assume that the dark matter is supersymmetric in nature .
given a supersymmetric model , the relic density of neutralinos is fixed and calculable in the standard cosmological model .
we use the relic density computed by edsj and gondolo @xcite , who include all neutralino self - annihilation diagrams , resonance and threshold effects , and chargino - neutralino coannihilations . in every model that is cosmologically allowed ,
the lightest neutralino is a cold relic from the big bang .
being a cold relic , we expect that it is attracted into the gravitational potentials of galaxies , and is at present a component of the cold dark halo .
we use the simple estimate that the dark matter density of the galactic halo at the position of the solar system is 0.3 gev @xmath42 .
neutralinos may make up the entire halo or only some fraction @xmath43 .
this fraction @xmath44 , where @xmath45 is the cosmological density of species @xmath46 measured in units of the critical density .
we now need an estimate of @xmath47 .
a common prescription @xcite is to take @xmath48 if @xmath49 , which is the value indicated by dynamical mass measurements of the amount of dark matter in galactic halos . if @xmath50 , this prescription sets @xmath51 .
this procedure is not conservative enough for our purposes because there may be more than one cold dark matter component in the galactic halo .
therefore , we use the constraint that @xmath52 , based on the age of the universe @xcite , to give a lower bound on @xmath53 . in this way
we obtain a conservative limit on the r - parity violating couplings .
this may be overly conservative , and indeed our final results would improve by approximately a factor of five if we used the less conservative fraction .
we consider neutralino decay through both the lepton - number violating term @xmath54 and the baryon - number violating term @xmath55 in the superpotential .
the coupling @xmath56 gives rise to the decays @xmath57 , @xmath58 and their charge conjugates .
the coupling @xmath25 gives rise to the decays @xmath59 ( @xmath60 ) and their charge conjugates .
the differential decay rates for these processes are calculated in ref .
@xcite . here
we quote the decay rate for @xmath61 in the limit where the squarks are heavy and degenerate , @xmath62 \nonumber,\end{aligned}\ ] ] where @xmath63 is the @xmath64 coupling constant , @xmath65 is the degenerate squark mass , and @xmath66 is the projection of the lightest neutralino on the interaction eigenstates @xmath67 , @xmath68 , @xmath69 , and @xmath70 .
when the gaugino fraction approaches unity , we see that the decay rate is @xmath71 when the gaugino fraction approaches zero , there is an additional suppression from the light quark masses , @xmath72 \nonumber.\end{aligned}\ ] ]
antiprotons are produced in the jets formed by the final state quarks .
we are interested in the energy spectrum of antiprotons @xmath73 at energies of a few gev .
it is quite difficult to calculate the spectrum @xmath73 , especially at low antiproton energies .
there are basically two theoretical frameworks available : a perturbative qcd approach and the lund monte - carlo . in fig .
[ fig : pbar ] we compare the theoretical curves with the measured proton spectra per quark jet in @xmath74 annihilations at various jet energies ( 14.5 gev @xcite , 29 gev @xcite , and 45.6 gev @xcite ) and also show the theoretical predictions at 1 tev , one of the highest energies we need . in the perturbative qcd approach , dokshitzer et al .
@xcite provide a formalism for calculating hadron spectra from parton spectra in the modified double - logarithmic approximation ( mlla ) .
the full mlla formula for inclusive particle production is too difficult to evaluate , and no numerical expression is available .
it has been shown that the simpler limiting spectrum reproduces the total charged particle spectrum quite well @xcite . however , as can be seen in fig .
[ fig : pbar ] , neither the limiting spectrum ( long - dashed curves @xcite ) nor the low energy limit of the full mlla spectrum ( short - dashed curves @xcite ) fit data for protons in a satisfactory manner . as a theoretical alternative to the mlla formalism , the lund monte carlo @xcite has been used by several authors , most recently by bottino et al .
however , it also does not match the data for protons satisfactorily ( histograms in fig .
[ fig : pbar ] ) . inspecting the experimental data and the results of the lund monte carlo , we observe that the ( anti)proton production rate in the high - energy tail of the spectrum increases with jet energy .
this is verified in the data for charged particle production in jets and is a generic prediction of the mlla formalism @xcite .
based on these arguments , we decide to take the value @xmath75 at the momentum @xmath76 gev in which we want the antiproton spectra ( see sect .
[ sec : res ] ) .
this underestimates the production rate of antiprotons , and gives a conservative bound on the r - parity violating couplings .
the number of jets per decay @xmath77 is equal to the number of final state quarks , except when there is a top quark .
the top decays before it can hadronize .
we approximate the process @xmath78 using the fact that the low energy proton spectrum is insensitive to the jet energy .
the bottom quark forms a jet .
in addition , we take the hadronic branching fraction of the @xmath79 , @xmath80 , as the probability to form two additional jets in the decay @xmath81 and its charge conjugate .
the final values for @xmath77 are 3 for a @xmath82 decay without top , 4.358 for a @xmath82 with top , 2 for an @xmath83 without top and 3.358 for an @xmath83 with top .
we now obtain the volume production rate of antiprotons by multiplying the flat antiproton spectrum by the number density of neutralinos @xmath84 and the number of jets per decay @xmath77 , and dividing by the lifetime @xmath85 , @xmath86
we use the diffusion model of webber _ et al . _ @xcite as implemented by chardonnet _
_ @xcite to describe diffusion of antiprotons through the galaxy .
the flux of antiprotons at the outer solar system is approximated by multiplying the source spectrum by the antiproton velocity and an appropriate diffusion time .
the diffusion constant in the galaxy depends on the rigidity @xmath87 of the particle and can be approximated by the following expression .
@xmath88 the region of turbulent magnetic fields responsible for diffusion can be approximated as a cylinder about 20kpc in radius and 2kpc high .
this gives a diffusion time for antiprotons of @xmath89 the flux at the outer solar system beyond the influence of the solar wind is now given by @xmath90 this flux is in units of antiprotons s@xmath91 @xmath92 gev@xmath91 sr@xmath91 .
the low energy spectrum is affected by the solar wind .
we use a simple model to account for this @xcite . the observed antiproton energy @xmath93 and momentum @xmath94
are given in terms of the initial energy outside the solar system @xmath95 .
@xmath96 we adopt values of the critical momentum @xmath97 and energy shift @xmath98 that correspond to the period of minimum activity in the 11 year solar cycle , @xmath991.105 gev and @xmath98=495 mev .
this gives the highest inner solar system flux , thus the conservative bound on the couplings . the solar modulated flux is now given by @xmath100
we proceed as follows .
we choose a data point for the @xmath101 flux at the earth @xcite .
these data have large error bars , but will improve with the next generation of experiments to be done in space @xcite . according to the data from bess , at an energy of 1.4 gev
the antiproton flux at the top of the atmosphere is 6.4 @xmath102^{-1}$ ] .
we denote this flux as @xmath103 . from the values of the observed energy @xmath93 and momentum @xmath94
, we determine the unmodulated @xmath101 energy @xmath95 .
this is the energy at which we evaluate the diffusion time @xmath104 and the velocity @xmath105 .
we now insist that the flux of antiprotons from neutralino decays be less than the observed flux . including corrections due to solar modulation , the statement is @xmath106 we choose not to make a correction for the spallation production of antiprotons .
the data are consistent with the picture that all cosmic ray antiprotons are made in spallation reactions , but there are a lot of uncertainties .
there is no good model for the amount we should subtract , and we have chosen to ignore it . we thus obtain a conservative bound on the couplings . using the data we have chosen ,
we obtain a model independent bound on the lifetime of a hadronically decaying relic with mass @xmath107 gev @xmath108 where we choose the mass fraction of neutralinos in the galactic dark halo as @xmath109 . since we assume that the neutralino is in the galactic halo at present , this bound applies for lifetimes longer than the age of the universe @xmath110 . in other words , we can exclude lifetimes in the range @xmath111 provided @xmath112 . for these long lifetimes ,
our bounds are in general better than those from diffuse photons @xcite , @xmath113 , and diffuse neutrinos @xcite , @xmath114 , except for some supersymmetric models with neutralino masses in the tev range .
we can limit each element of the two coupling matrices @xmath23 and @xmath24 separately , by choosing the coupling such that all of the measured @xmath101 flux comes from the neutralino decay via a single channel .
our excluded range of lifetimes then translates into the following _ excluded _ range of @xmath115 @xmath116 where @xmath117 is the upper limit on @xmath115 coming from the antiproton flux and @xmath118 is the value of @xmath115 for which the neutralino lifetime equals the age of the universe , which sets the validity of our analysis . in the previous formulas
, @xmath119 is in gev , @xmath120 is in seconds , and the age of the universe is taken to be @xmath121 yr . for virtually all models we find that @xmath122 is smaller than @xmath123 by at least three orders of magnitude . conservatively taking @xmath124
we find for the excluded range that @xmath125 if we would further demand that the neutralino be cosmologically interesting , namely @xmath126 , we would expand the excluded range to @xmath127 .
we have plotted the upper bound on @xmath128 as a function of the neutralino mass for 10@xmath129 supersymmetric models in figure [ fig : llmx ] .
the results are similar for all generations in the r - parity violating couplings , except for @xmath130 which shows the top quark threshold ( see figure [ fig : lltop ] ) .
when the neutralino is lighter than the top quark , there is no bound on @xmath130 at tree level .
a bound can however be found at the one loop level @xcite . in our sample of supersymmetric models , we find the absolute upper limits @xmath131 is due to our neglect of one - loop diagrams .
these upper limits depend on the neutralino mass , and improve considerably with higher masses .
we find the following power law fits @xmath132 for comparison , the most stringent laboratory constraints , those from proton decay @xcite , bound not @xmath133 and @xmath134 separately but their product to @xmath135 , where @xmath26 is the sfermion mass scale .
these bounds are six orders of magnitude weaker than ours . on the other hand , because our analysis only applies to lifetimes longer than the age of the universe , our excluded region spans only 3 orders of magnitude .
as mentioned in the introduction , the remaining gap of 3 orders of magnitude may be filled by other cosmological considerations .
we finally comment on the dependence of our bounds on model parameters . in figure
[ fig : llmsq ] we plot the dependence on the squark mass scale @xmath26 .
as expected , the bound is better for lighter squarks . regarding the dependence on neutralino composition , in general the bounds are better in the gaugino region .
in the higgsino region , the bounds improve with heavier generations as the yukawa couplings become larger .
the dependence on other model parameters is negligible in comparison .
from observations of cosmic ray antiprotons we obtain strong constraints on each of the hadronic r - parity violating couplings @xmath133 and @xmath134 that appear in extensions of the minimal supersymmetric model .
our analysis applies strictly in cases where the neutralino lifetime is longer than the age of the universe , and enables us to exclude a range @xmath136 , where @xmath122 is the upper limit from the antiproton flux and @xmath123 is the validity boundary coming from the age of the universe . using a large selection of model parameters , we find absolute upper limits @xmath137 and @xmath138 when there is no top quark in the final state , and @xmath139 when there is . in
virtually all models , we find that @xmath140 .
our bounds lie many orders of magnitude below the laboratory constraints coming from proton decay , and apply to each r - parity violating coupling separately .
we thank j. silk for useful discussions and comments on the manuscript .
p.g . thanks him also for the friendly hospitality and encouraging support at the center for particle astrophysics .
p.g . thanks s. lupia for useful discussions and programs on the theoretical proton spectra .
this research was supported in part by nasa grant 1 - 443839 - 23254 - 2 and doe grant fg03 - 84er40161 .
g. cowan ( aleph collaboration ) , talk presented at the special cern particle physics seminar on physics results from the lep run at 172 gev , 25 february , 1997 , http://alephwww.cern.ch/alpub/seminar/cowan-172-jam/cowan.html .
l. dokshitzer , v. a. khoze , and s. i. troyan , int . j. mod
a * 7 * , 1875 ( 1996 ) ; yu .
l. dokshitzer , v. a. khoze , a. h. mueller , and s. i. troyan , _ basics of perturbative qcd _
( editions frontires , 1991 ) .
s. lupia and w. ochs , phys .
b * 365 * , 339 ( 1996 ) ; v. a. khoze , s. lupia , and w. ochs , _ ibid .
_ * 394 * 179 ( 1997 ) ; w. ochs , in _ ichep 96 : proceedings , july 1996 , warsaw , poland _ , edited by z. ajduk and a. k. wroblewski ( world scientific , 1997 ) ,
p. 25 - 31 ; s. lupia , to appear in _ proceedings of the 33rd eloiastron workshop `` universality features in multihadron production and the leading effect '' _ , october 1996 , erice , italy , preprint ( 1997 ) . | we constrain the hadronic r - parity violating couplings in extensions to the minimal supersymmetric standard model .
these interactions violate baryon and lepton number , and allow the lightest superpartner ( lsp ) to decay into standard model particles .
the observed flux of cosmic ray antiprotons places a strong bound on the lifetime of the lsp in models where the lifetime is longer than the age of the universe .
we exclude @xmath0 and @xmath1 except in the case of a top quark , where we can only exclude @xmath2 . |
let @xmath2 be a finite graph .
a @xmath3-_coloring _ of @xmath2 is a surjective function @xmath4 that assigns a number from the set @xmath5:=\{1,\dots , k\}$ ] to each vertex of @xmath2 .
a @xmath3-coloring @xmath4 of @xmath2 is called _ proper _ if any two adjacent vertices have different colors , and @xmath4 is called _ complete _ if for each pair of different colors @xmath6 $ ] there exists an edge @xmath7 such that @xmath8 and @xmath9 .
a @xmath3-coloring @xmath4 of a connected graph @xmath2 is called _ connected _ if for all @xmath10 $ ] , each color class @xmath11 induces a connected subgraph of @xmath2 .
the _ chromatic number @xmath12 of @xmath2 _ is the smallest number @xmath3 for which there exists a proper @xmath3-coloring of @xmath2 .
the _ hadwiger number _
@xmath13 is the maximum @xmath3 for which a connected and complete coloring of a connected graph @xmath2 exists , and it is defined as the maximum @xmath14 among the connected components @xmath0 of a disconnected graph @xmath2 ( it is also known as _ the connected - pseudoachromatic number _ , see @xcite ) .
a graph @xmath0 is called a _
minor _ of the graph @xmath2 if and only if @xmath0 can be formed from @xmath2 by deleting edges and vertices and by contracting edges .
suppose that @xmath15 is a minor of a connected graph @xmath2 .
if @xmath16 $ ] then there exists a natural corresponding complete @xmath3-coloring @xmath17 $ ] for which @xmath11 is exactly the set of vertices of @xmath2 which contract to vertex @xmath18 in @xmath15 .
the _ hadwiger number _ @xmath13 of a graph @xmath2 is the largest @xmath3 for which @xmath15 is a minor of @xmath2 . clearly ,
@xmath19 where @xmath20 denotes the _ clique number _ of @xmath2 : the maximum clique order of @xmath2 .
the hadwiger number was introduced by hadwiger in 1943 @xcite together the _ hadwiger conjecture _ which states that @xmath21 for any graph @xmath2 .
the following definition is an extension of the notion of _ perfect graph _ , introduced by berge @xcite :
let @xmath22 be two distinct parameters of @xmath2 .
a graph @xmath23 is called _ @xmath24-perfect _ if for every induced subgraph @xmath25 of @xmath23 , @xmath26 . note that , with this definition a perfect graph is denoted by @xmath27-perfect .
the concept of the @xmath28-perfect graphs was introduced by christen and selkow in @xcite and extended in @xcite .
a graph @xmath2 without an induced subgraph @xmath0 is called @xmath0-_free_. a graph @xmath29-free , @xmath30-free , ... is called _ @xmath31-free_. a _ chordal graph _ is a @xmath32-free one . some known results are the following : l ' ovasz proved in @xcite that a graph @xmath2 is @xmath33-perfect if and only if its complement is @xmath33-perfect .
chudnovsky , robertson , seymour and thomas proved in @xcite that a graph @xmath23 is @xmath27-perfect if and only if @xmath23 and its complement are @xmath34-free . this paper is organized as follows : in section [ 2 ] we prove that the families of chordal graphs and the family of @xmath35-perfect graphs are the same . in section
[ 3 ] , we give some consequences of the section [ 2 ] as characterizations of other graph families related to complete colorings .
we will use the following chordal graph characterization to prove theorem [ teo1 ] : [ hajnal , sur ' anyi @xcite and dirac @xcite ] [ teo0 ] a graph @xmath2 is chordal if and only if @xmath2 can be obtained by identifying two complete subgraphs of the same order in two chordal graphs .
now , we characterize the chordal graphs and the @xmath35-perfect ones . the following proof is based on the standard proof of the chordal graph perfection ( see @xcite ) .
[ teo1 ] a graph @xmath2 is @xmath35-perfect if and only if @xmath2 is chordal .
assume that @xmath2 is @xmath35-perfect .
note that if a cycle @xmath0 is one of four or more vertices then @xmath36 and @xmath37 .
hence , every induced cycle of @xmath2 has at the most @xmath38 vertices and the implication is true .
now , we verify the converse . since every induced subgraph of a chordal graph is also a chordal graph , it suffices to show that if @xmath2 is a connected chordal graph , then @xmath39 . we proceed by induction on the order @xmath40 of @xmath2 . if @xmath41 , then @xmath42 and @xmath43 .
assume , therefore , that @xmath44 for every induced chordal graph @xmath0 of order less than @xmath40 for @xmath45 and let @xmath2 be a chordal graph of order @xmath40 . if @xmath2 is a complete graph , then @xmath46 .
hence , we may assume that @xmath2 is not complete .
by theorem [ teo0 ] , @xmath2 can be obtained from two chordal graphs @xmath29 and @xmath30 by identifying two complete subgraphs of the same order in @xmath29 and @xmath30 .
let @xmath47 denote the set of vertices in @xmath2 that belong to @xmath29 and @xmath30 .
thus the induced subgraph @xmath48 in @xmath2 by @xmath47 is complete and no vertex in @xmath49 is adjacent to a vertex in @xmath50 .
hence , @xmath51 moreover , according to the induction hypothesis , @xmath52 and @xmath53 , then @xmath54 on the other hand , since @xmath47 is a clique cut then each walk between @xmath49 and @xmath50 contains at least one vertex in @xmath47 .
let @xmath4 be a pseudo - connected @xmath13-coloring of @xmath2 , and suppose there exist two color classes such that one is completely contained in @xmath49 , and the other one is completely contained in @xmath50 .
clearly these two color classes do not intersect , which contradicts our choice of @xmath4
. moreover , each color class with vertices both in @xmath49 and in @xmath50 , contains vertices in @xmath47 .
consequently , every pair of color classes having vertices both in @xmath49 and in @xmath50 must have an incidence in @xmath55 .
thus , @xmath56 by equation [ des1 ] , @xmath57 and the result follows .
it is known that every chordal graph is a @xmath33-perfect one ( see @xcite ) .
the following corollary is a consequence of the chordal graph perfection .
every @xmath35-perfect graph is @xmath33-perfect .
in this section , we give a new characterization of several family of @xmath24-perfect graphs related to complete colorings .
complete bipartite graphs have achromatic number two ( see @xcite ) but their hadwiger number can be arbitrarily large , while the graph formed by the union of @xmath62 has hadwiger number two but its achromatic number can be arbitrarily large . therefore , @xmath63 and @xmath64 are two non comparable parameters .
we will use the following characterization in the proof of corollary [ cor1 ] . on one hand , a coloring such that every color class contains a vertex that has a neighbor in every other color class is called _ dominating_. the _
pseudo-@xmath1-chromatic number _
@xmath72 of a graph @xmath2 is the largest integer @xmath3 such that @xmath2 admits a dominating @xmath3-coloring .
[ cor3 ] for any graph @xmath23 the following are equivalent : @xmath75 @xmath2 is @xmath65-perfect , @xmath76 @xmath2 is @xmath77-perfect , @xmath78 @xmath2 is @xmath79-perfect and @xmath80 @xmath2 is @xmath81-free
. the proofs of @xmath82 and @xmath83 immediately follow from ( [ des5 ] ) . to prove @xmath84 note that , if @xmath85 then @xmath86
, hence the implication is true , see figure [ fig1 ] .
the proof of @xmath87 is a consequence of theorem [ teo2 ] .
first , a coloring of @xmath2 is called _ pseudo - grundy _ if each vertex is adjacent to some vertex of each smaller color . the _ pseudo - grundy number _
@xmath91 is the maximum @xmath3 for which a pseudo - grundy @xmath3-coloring of @xmath2 exists ( see @xcite ) .
second , a proper pseudo - grundy coloring of @xmath2 is called _ grundy_. the _ grundy number _ @xmath92 ( also known as the _ first - fit chromatic number _ ) is the maximum @xmath3 for which a grundy @xmath3-coloring of @xmath2 exists ( see @xcite ) . from the definitions , we have that for the converse , let @xmath2 be a @xmath94-perfect graph . if @xmath0 is an induced graph of @xmath2 , by equation [ des4 ] , @xmath100 . since @xmath2 is a chordal graph , @xmath44
, so the implication follows . for the converse , a @xmath105-free graph @xmath2 is @xmath94-perfect ( by theorem [ teo3 ] ) and @xmath103-perfect ( by theorem [ teo4 ] ) .
then , for every induced subgraph @xmath0 of @xmath2 , @xmath108 by equation [ des4 ] and @xmath109
. therefore , @xmath110 and the result follows .
r. p. gupta , _
bounds on the chromatic and achromatic numbers of complementary graphs .
_ , recent progress in combinatorics ( proc .
third waterloo conf . on comb . , 1968 ) , academic press , new york , 1969 , pp . | a graph is _ chordal _ if every induced cycle has three vertices .
the _ hadwiger number _ is the order of the largest complete minor of a graph .
we characterize the chordal graphs in terms of the hadwiger number and we also characterize the families of graphs such that for each induced subgraph @xmath0 , ( 1 ) the hadwiger number of @xmath0 is equal to the maximum clique order of @xmath0 , ( 2 ) the hadwiger number of @xmath0 is equal to the achromatic number of @xmath0 , ( 3 ) the @xmath1-chromatic number is equal to the pseudoachromatic number , ( 4 ) the pseudo-@xmath1-chromatic number is equal to the pseudoachromatic number , ( 5 ) the hadwiger number of @xmath0 is equal to the grundy number of @xmath0 , and ( 6 ) the @xmath1-chromatic number is equal to the pseudo - grundy number . * _ keywords : _ * complete colorings , perfect graphs , forbidden graphs characterization . *
2010 mathematics subject classification : * 05c17 ; 05c15 ; 05c83 . |
is a high mass x - ray binary associated with the galactic plane variable radio source @xcite which shows periodic non - thermal outbursts on average every @[email protected] d @xcite .
optical spectroscopic observations show that the system is composed of a rapidly rotating early type b0ve star with a stable equatorial decretion disk and mass loss , and a compact object with a mass between 1 and 4 @xmath9 orbiting it every @xmath626.5 d @xcite .
spectral line radio observations give a distance of [email protected] kpc @xcite .
@xcite reported the discovery of an extended jet - like and apparently precessing radio emitting structure at angular extensions of 1050 milliarcseconds . due to
the presence of ( apparently relativistic ) radio emitting jets , was proposed to be a microquasar .
however , recent vlba images obtained during a full orbital cycle show a rotating elongated morphology @xcite , which may be consistent with a model based on the interaction between the relativistic wind of a non - accreting pulsar and the wind of the stellar companion @xcite .
the magic cerenkov telescope has recently detected at very high energy gamma rays ( @xmath10100 gev ; @xcite ) . and
@xcite share the quality of being the only two known x - ray binaries that are also gev emitters .
therefore , they can be called gamma - ray binaries . has been observed with different x - ray satellites : _
@xcite , _ rosat _
@xcite , _ asca _
@xcite , _ rxte _
@xcite , _ bepposax _ , _ xmm - newton _ , and _ integral _ @xcite . the spectrum could always be fitted with an absorbed power - law with values in the ranges @xmath11=0.450.65@xmath12 @xmath2 and @xmath13=1.51.9 .
the 210 kev x - ray flux shows orbital variability @xcite , with values ranging from about 5 to 20@xmath14 ergs @xmath2 s@xmath3 , the maximum occurring at orbital phase @xmath60.5 ( assuming @xmath15=jd 2,443,366.775 ) .
no spectral lines nor extended x - ray emission have ever been reported . here
we present , for the first time , _ chandra _ x - ray observations of , aimed at detecting small - scale extended x - ray emission .
we observed ls i + 61 303 with _
chandra _ using the standard acis - i setup in vf mode and 3.24 s time frame during a total of 49.9 ks from 2006 april 7 22:30 ut to april 8 12:22 ut ( mjd 53832.937553833.5153 ) .
this corresponds to orbital phases 0.0280.050 , when a relatively low flux is expected , thus minimizing possible pileup effects .
the _ chandra _ interactive analysis of observations software package ( ciao 3.3.0.1 ) and the caldb 3.2.2 version have been used to perform the reduction of the level 1 event files as well as for the spectrum and lightcurve subtraction . the reprocessed level 2 event files
have been created to account for readout effects .
the position angle of the readout direction of the acis - i chip was @xmath16 ( positive from north to east ) .
we have reduced the data using standard procedures given in the _ chandra _ analysis threads .
the source region has been defined as a 20 pixel ( @xmath610 ) radius circle around the source center . for the background
, we took a nearby circular region of 80 pixel ( @xmath640 ) radius , away from the source and the ccd junction . to avoid the presence of possible ( very faint ) sources in the background ,
the tool _ celldetect _ was applied .
the _ mkacisrmf _ and _ mkarf _ tools were used to generate the rmf and arf files respectively .
the tool _ dmextract _ was used to generate the source and the background lightcurves and spectra .
the spectrum was grouped with the tool _ dmgroup _ to 30 counts per energy bin
. showed a moderate level of activity , with an average count rate of 0.15 counts s@xmath3 ( see fig .
[ fig : lc ] ) , and the data was affected by pileup at the @xmath614% level ( see below ) .
the background was very low and estimated to be about 1% of the source count rate within the source region .
the position of the x - ray source , determined by using the ciao package tool _ celldetect _ , is @xmath17 , @xmath18 ( j2000 ) .
the errors , as in the rest of this work , are 1@xmath0 .
the x - ray position is consistent with the radio and optical positions .
the background subtracted lightcurve of in the 0.310 kev band is plotted at the bottom of fig .
[ fig : lc ] .
each point represents 324 s of data and the error bars represent the square root of the total number of counts in each bin divided by 324 s. we have also computed the hardness ratio ( hr ) by dividing the count rate in the 1.710 kev energy range by the corresponding one in the 0.31.7 kev range .
we show at the top of fig .
[ fig : lc ] the hr as a function of time , but with a larger bin time of 1024 s to have relatively small error bars .
the count rate is moderately variable on timescales from several minutes to hours , with an average of [email protected] counts s@xmath3 ( @xmath19=1.84 ) .
the count rate increases during the observation , and a linear fit yields a 30% increase at the end with respect to the beginning .
superimposed on this trend is a miniflare 32 ks after the start of the observation , when the count rate increases by a factor of @xmath62 over a timescale of @xmath61000 s , with a total duration of @xmath63 ks .
this miniflare is accompanied by an increase of the hr .
this kind of moderate variability was first observed with _ asca _ , which detected source variations of about a 50% on timescales of half an hour @xcite .
archival _ bepposax _ observations carried out in 1997 have also shown short - term variability @xcite .
it is interesting to note that small amplitude ( @xmath2010% ) radio flares were detected in with a recurrence period of about 5 ks @xcite .
the presence of hour timescale x - ray miniflares has also been reported in the case of @xcite . a hardness vs. intensity diagram for the whole _
observation of reveals a positive trend , the source is harder when it is brighter , but with a low significant correlation coefficient of @xmath21=0.44 .
this is in contrast to the clear correlation found in previous _
observations by @xcite . since the hr is not significantly variable ( @xmath19=1.78 ) , we considered the whole data set for the spectral analysis . to perform
the spectral fits we restricted the energy range between 0.3 and 10 kev .
a fit to the data with a bremsstrahlung model provided @xmath19=1.6 , and no lines nor a multicolor blackbody were required to fit the data .
an absorbed power - law fit yielded @xmath19=1.14 , but a significant excess above 7 kev was apparent in the residuals .
due to the count rate and time frame used , this is probably due to pileup .
the addition of the pileup model implemented in _ sherpa _
@xcite to the absorbed power law yielded different sets of possible spectral parameter values , all of them equally compatible from a statistical point of view .
nevertheless , for some of them the derived fluxes were a factor @xmath225 higher than previous measurements at similar orbital phases .
we found that reasonable results were obtained when fixing the event pileup fraction parameter @xmath23 to 0.95 : @xmath19=0.83 , grade migration parameter @xmath24 , @xmath25 @xmath2 , @xmath5 , pileup fraction 14% , and @xmath1 ergs @xmath2 s@xmath3 .
the significance of adding a pileup component to the absorbed powerlaw fit was estimated via an f - test : the probability that this improvement happens by chance is 0.8% . to check the robustness of the results obtained with the spectral fitting with the pileup model we have conducted two additional tests .
first of all we have obtained the spectrum of the readout streak , which is not affected by pileup , after excluding the 5 pixel diameter region around the source .
since it only contains 287 counts , we have fitted the spectrum using low count number statistical weights .
fixing @xmath11 to @xmath26 @xmath2 we have obtained @xmath27 , which is consistent within errors to the value given above .
an additional test has been conducted by fitting the spectrum of the source wings , which are not affected by pileup .
we have built a spectrum with data beyond a radius of 1.5 from the source center ( i.e. , excluding the inner @xmath28 pixels , those most affected by pileup ) and fitted it with and without fixing @xmath11 , obtaining in both cases a photon index harder than the one obtained with the pileup model .
a similar test with data beyond 2.0 has provided similar results .
these tests clearly suggest that the photon index value obtained with the pileup model is reliable , and that the source was intrinsically hard during this observation .
moreover , the estimate of the total source flux using the fitted flux of the source wings and the fraction of the simulated psf ( see next section ) counts contained in the wings is compatible for @xmath23=0.95 .
this is not the case for the lower values of @xmath23 that implied the much higher source fluxes discussed in the previous paragraph .
these results give us confidence on the results obtained with the pileup model .
regarding the emission of itself , the flux we report here , @xmath1
ergs @xmath2 s@xmath3 , is compatible within errors to that obtained during a short _ xmm _ observation also around phase 0 @xcite .
however , the photon index @xmath5 found in the _ chandra _ data is the hardest ever found , only compatible with any previous value at the @xmath29 level and harder , at a @xmath65@xmath0 level , than the _ xmm _ one at the same phase ( @xmath30 ) . it is outside the scope of this work to put forward a physical model for the spectral state of the source .
we will just point out that the x - ray photon index can be harder than the canonical value of 1.5 , which corresponds to a electron population with power - law index of 2 .
this can occur for different reasons , for example a klein - nishina dominated steady particle population ( see @xcite for an extended discussion ) .
a very hard spectral component ( e.g. inverse compton or bremsstrahlung ) , produced by the lower energy part of the particle spectrum with a high low - energy cutoff , may also harden the synchrotron spectrum .
the lack of simultaneous data at higher energies does not allow us to distinguish between the two hypotheses .
we finally selected the data corresponding to the miniflare and obtained a spectrum for it , which had poor statistics and was especially affected by pileup . a fit to this spectrum yielded a photon index compatible within ( large ) errors to that of the whole observation spectrum .
a preliminary visual inspection of the source image reveals the possible existence of extended emission .
to better study this issue we have used _
chart _ and _ marx _ to simulate the point spread function ( psf ) accounting for pileup effects .
this psf has the same counts as the source , which are distributed in energy according to the source spectrum including pileup , being spatially distributed in the image considering pileup as well .
we show in fig .
[ fig : images ] the images corresponding to the source , the psf , and the residuals obtained after subtracting the psf from the source , all of them smoothed with a gaussian kernel of three pixels in radius to enhance the faint emission . as can be seen in fig .
[ fig : images]-left , there appears to be extended x - ray emission around at distances between 5 and 12 from the source center .
the subtraction of the nearly circular psf , shown in fig .
[ fig : images]-middle , from the source image yields the residuals shown in fig .
[ fig : images]-right .
there seems to be a region with an excess of counts at several arcseconds from the image core and nearly half - surrounding it .
this excess region , already apparent in the source image , presents some spots and extends mainly towards the north - east . to check the reliability of the 512 excess we have created radial profiles of the surface brightness for the source and the simulated psf with the tools _ dmextract _ and _ dmtcalc_. we show the radial profiles with bins of 1 in fig .
[ fig : radial ] , where hints of excess are seen in the region 512 away from the center .
moreover , we note that a comparison with the unscattered point - like blazar source located at high - galactic latitude ( following the procedure described in @xcite ) provided an even clearer departure of from a point - like source . nevertheless , we prefer to show here the most conservative results obtained with the psf , simulated at the chip position and with the observed spectrum of including pileup effects .
next we computed the excess of source counts over the psf in an annulus with inner and outer radii of 5 and 125 , and obtained 58@xmath818 counts ( 3.2@xmath0 , or single trial probability of @xmath31 ) .
we have looked for excesses in a circle with radius 20 centered on the source , without considering the central 5 radius circle , thus yielding an area of @xmath32@xmath33 . dividing this by the area of the annulus where we obtained the 3.2@xmath0 excess we find that the number of trials is 2.9 , yielding a probability of @xmath34 , which corresponds to 2.9@xmath0 .
however , since the extended emission appears to be asymmetric , we have computed the excess dividing the 5125 annulus in four quadrants .
although there is no excess towards the west and south directions , the excess is of about 1.5@xmath0 to the east , and reaches 3.8@xmath0 to the north ( or single trial probability of @xmath35 ) .
we note that this excess is not located in the path of the readout streak .
the number of trials is now 11.6 , providing a post - trial probability of @xmath36 or a 3.2@xmath0 detection .
the excess in the northern quadrant of the 5125 annulus , with a projected distance of 0.050.12 pc from the center of , represents 0.5% of the source counts , implying ( with @xmath13=1.25 ) a flux of @xmath64@xmath37 ergs @xmath2 s@xmath3 and @xmath382@xmath39 ergs s@xmath3 .
the chandra observations reported here show that presented moderate fluxes of @xmath40 ergs @xmath2 s@xmath3 , as expected around orbital phase @xmath41 ( @xcite ) .
the source appeared to be moderately variable on timescales of @xmath61000 s. the hydrogen column density was @xmath25 @xmath2 , similar to the values found in the past .
the photon index was @xmath42 , harder than the previously measured values in the range 1.51.9 .
finally , an excess between 5 and 125 towards the north of has been detected at the 3.2@xmath0 level ( post - trial ) .
_ chandra _ observations of x - ray binaries have provided the most detailed x - ray images of these sources . extended circular emission , produced by ism dust grains along the line of sight that scatter mainly the soft x
- rays @xcite , has been found in most cases @xcite .
however , the possible extended structure around we report here is asymmetric , suggesting its intrinsic nature .
plausible explanations of this diffuse radiation could be thermal bremsstrahlung produced by hot and dense ism close to the source , perhaps heated and compressed by a large - scale outflow originated in , and synchrotron or inverse compton emission produced by non - thermal particles accelerated at the termination region of the aforementioned outflow .
the first possibility would require a medium with densities significantly higher than those inferred from the ism absorption towards .
if non - thermal , the extended emission would more likely be of synchrotron origin due to its shorter timescales to radiate at a few kev than those of inverse compton or relativistic bremsstrahlung . in any case , the asymmetric distribution of counts could be related to the main direction of the outflow and/or to ism inhomogeneities .
regardless of the morphology , we can say that efficient energy transport is probably taking place up to distances of at least @xmath60.1 pc in .
however , only deeper high - resolution observations can help to better detect the excess ( currently at a 3.8@xmath0 level , 3.2@xmath0 post - trial ) and determine the precise morphology and underlying physics .
a final comment on the hydrogen column density appears necessary .
the @xmath43 value inferred from spectral fits to all available x - ray observations is in the range 0.450.70@xmath44 @xmath2 , while the @xmath11 value of the ism inferred from uv / optical absorption is in the range 0.450.60@xmath45 @xmath2 .
therefore , the detected x - rays suffer small intrinsic absorption , if any .
moreover no significant @xmath11 orbital variations have ever been detected ( as pointed out previously by @xcite after _ asca _ observations ) . in conclusion ,
the apparent lack of significant soft x - ray absorption and the constancy of @xmath11 along the orbit are difficult to be explained considering the slow and dense material expelled from the companion star .
the x - ray emitter should be placed away from the binary system to produce such observational results , in contrast with the expectations from classical colliding wind and microquasar scenarios .
this issue certainly requires further investigation .
+ we thank mike nowak for helpful discussions on the analysis of the _ chandra _ data .
, v.b - r . and j.m .
acknowledge support by dgi of the spanish ministerio de educacin y ciencia ( mec ) under grants aya2004 - 07171-c02 - 01 and aya2004 - 07171-c02 - 02 and feder funds .
m.r . acknowledges financial support from mec through a _
juan de la cierva _ fellowship .
thanks the max - planck - institut fr kernphysik for its support and kind hospitality .
j.r.w . was supported by cycle 7 chandra grant go6 - 7025x .
d.f.t . has been supported by mec under grant aya-2006 - 00530 , and the guggenheim foundation .
is also supported by pai of junta de andaluca as research group fqm322 . | we present a 50 ks observation of the gamma - ray binary carried out with the acis - i array aboard the _ chandra _ x - ray observatory .
this is the highest resolution x - ray observation of the source conducted so far .
possible evidence of an extended structure at a distance between 5 and 12 towards the north of have been found at a significance level of 3.2@xmath0 .
the asymmetry of the extended emission excludes an interpretation in the context of a dust - scattered halo , suggesting an intrinsic nature . on the other hand , while the obtained source flux , of @xmath1 ergs @xmath2 s@xmath3 , and hydrogen column density , @xmath4 @xmath2 , are compatible with previous results , the photon index @xmath5 is the hardest ever found . in light of these new results , we briefly discuss the physics behind the x - ray emission , the location of the emitter , and the possible origin of the extended emission @xmath60.1 pc away from . |
the optical color - magnitude relation of galaxies shows a clear bimodality , with a narrow sequence of red ( predominantly spheroidal ) galaxies and a `` blue cloud '' of actively star - forming galaxies ( e.g. , @xcite , @xcite ) .
studies of the evolution of the red galaxy luminosity function suggest that the stellar mass within the red sequence has increased since @xmath7 , perhaps by a factor of @xmath8 ( @xcite ; @xcite ; @xcite ) .
the most likely cause of this growth is the transition of blue cloud galaxies to red sequence galaxies at @xmath9 .
recent semi - analytic models of galaxy formation support such a scenario ( e.g. , @xcite ) . in order to reproduce the color bimodality ,
one or more mechanisms must be introduced to shut down the star - formation ( e.g. , @xcite ) , and turn blue galaxies red .
some authors suggest that feedback from merger - induced agn activity could be the quenching mechanism ( e.g. , @xcite , @xcite ) , while others suggest that supernovae in a starburst galaxy may have the required energy to blow out enough gas @xcite .
other possible mechanisms include virial shock heating of the galaxy halo @xcite or simple gas consumption .
if we are to understand how galaxies shut off their star - formation and become red , it is clearly important to identify and study the population of sources that have recently moved onto the red sequence or are in a transition region between the red sequence and blue cloud commonly referred to as the `` green valley '' .
these sources may still show signs of agn and/or residual star - formation as they move onto the red sequence .
optical studies of the stellar populations of red galaxies generally show them to be dominated by an old component ( e.g. , @xcite ) , with perhaps a `` frosting '' of stars formed since @xmath10=1 ( e.g. , @xcite ) . in studies of local ellipticals
, the weak infrared emission is usually attributed to mass loss from the evolved stellar population ( @xcite ; @xcite ) . however , there has been recent evidence from @xmath11 and @xmath4 that both hidden agn and star - formation activity may occur in red galaxies out to @xmath12 ( e.g. , @xcite , @xcite , @xcite ) .
more than half of all red galaxies have narrow emission lines in their optical spectra indicative of agn activity ( e.g. , @xcite ; @xcite ) .
we have identified a population of red galaxies with relatively strong 24@xmath1 emission ( [email protected] mjy ; hereafter `` 24@xmath1 excess '' red galaxies ) .
the redshift range [email protected] allows us to use both optical and infrared diagnostics to determine whether they are likely to be dominated by agn or star - formation activity .
we use starburst templates @xcite to estimate the total infrared luminosities of l@xmath15 at @xmath10 = 0.15 ( 0.3 ) .
assuming that all the dust heating is due exclusively to star formation processes and adopting the @xcite imf ( 0.1 - 100 @xmath16 ) , this corresponds to sfrs of @xmath22 ( 5 ) m@xmath17 @xcite .
this infrared luminosity is orders of magnitude larger than expected from local ellipticals ( which have infrared luminosities from 5@xmath18 - 7@xmath19 ; @xcite ) and suggests optically obscured activity in these galaxies .
similar results are found by @xcite for @xmath20 spheroidal galaxies in goods @xcite and by @xcite for z@xmath21 red galaxies in swire @xcite . in this paper
, we investigate the nature of the 24@xmath1 excess in these galaxies .
we use mid - ir colors to determine whether the sources have significant polycyclic aromatic hydrocarbon ( pah ) emission features that are known to be strong in luminous star - forming galaxies .
we also use optical line diagnostics to distinguish between photo - ionization by young massive stars in star - forming hii regions and by agn .
the nature of the 24@xmath1 emission from optically selected agn and other populations will be explored further in w. rujopakarn in preparation .
a cosmology of @xmath22 , @xmath23=0.3 , and @xmath24=0.7 is assumed throughout .
all optical magnitudes are on the vega system .
our sample is selected from the agn and galaxy evolution survey ( ages ; c. s. kochanek et al . in prep . ; j. moustakas et al . in prep . ) .
ages is an optical spectroscopic survey conducted within the botes field of the noao deep wide - field survey ( ndwfs ; @xcite ) .
the survey was carried out in 2004 and 2005 using the hectospec multi - fiber spectrograph ( @xcite ; @xcite ) at the mmt 6.5 m telescope .
successful redshifts were measured for nearly 20,000 objects . for our analysis , we use the ages `` main galaxy sample '' , which is statistically complete for @xmath25 , after correcting for sparse sampling , fiber incompleteness , and redshift failures ( see eisenstein et al .
, in preparation for details ) .
the ages sparse sampling preferentially targets brighter optical galaxies and/or sources detected in other wavelengths and results in roughly half of the total @xmath2619.95 sample with good quality optical spectra .
we account for any selection effects in which one type of source is selected more frequently than another by applying a correcting weight to each galaxy ( each individual galaxy is weighted by the inverse of the known sampling rate of galaxies with similar properties ) .
the final percentages of red , green , blue and 24 @xmath1 excess red galaxies with different optical and infrared classifications only changes by a few percent at the most .
the optical spectroscopy covers the wavelength range 3700 - 9200 at a fwhm resolution of @xmath26 .
the ages spectra were reduced using standard procedures implemented in @xmath27 , a customized hectospec reduction package .
the rest - frame optical emission lines were modeled using gaussian line profiles , after carefully subtracting the underlying stellar continuum using the @xcite population synthesis models ; additional details regarding the ages data reduction and emission line measurements can be found in j. moustakas et al . in preparation .
the botes field covers a contiguous 9.3 square degree area and has a plethora of existing multi - wavelength imaging and spectroscopy .
the field has been mapped in @xmath28 , @xmath29 , and @xmath30 bands to median 3@xmath31 point source depths of @xmath3227.7 , 26.7 , and 26.0 ( vega ) respectively .
the _ spitzer space telescope _
@xcite has imaged the field at 24 , 70 , and 160 @xmath1 using the multiband imaging photometer for @xmath4 ( mips ; @xcite ) to 5@xmath31 rms depths of 0.3 mjy , 25 mjy , and 150 mjy respectively , yielding a catalog of @xmath3222,000 sources .
@xmath4 has also imaged the field with the infrared array camera ( irac ; @xcite ) to 5@xmath31 depths of 6.4 , 8.8 , 51 , and 50 @xmath33jy at 3.6 , 4.5 , 5.8 , and 8@xmath1 respectively @xcite .
we have selected a sample of 3889 galaxies from the ages survey that have spectroscopic redshifts , @xmath34 and absolute v - band magnitudes of m@xmath35 - 19.4 mag .
this redshift range allows the irac color - color diagram to be used as a diagnostic for the presence of pah emission ( with the strong pah emission features redshifted into the observed 8@xmath1 band ) and ensures that we cover the rest - frame wavelengths of several powerful emission line diagnostics , including h@xmath36 , h@xmath37 , [ oiii ] @xmath384959,5007 , and [ nii ] @xmath386548,6584 . at absolute magnitudes
fainter that m@xmath39 - 19.4 mag , the number of red galaxies decreases significantly .
our absolute optical magnitude cut ensures that we are comparing red and blue galaxies within a similar absolute magnitude range .
the optical magnitude cut also ensures that these are galaxies selected from the main ages sample ( with @xmath2619.95 ) that has easily quantifiable selection effects .
the rest - frame optical luminosities and colors were computed using @xmath40 ( v. 4.1.4 ; @xcite ; @xcite ) . for more details ,
the reader is referred to j. moustakas et al . in preparation .
we classify the galaxies into `` red sequence '' ( 914 sources ) and `` blue cloud '' ( 2255 sources ) using the following rest - frame selection criteria : + red sequence : @xmath41 blue cloud : @xmath42 the 720 sources that fall between these criteria on the color - magnitude diagram are classified as `` green valley '' sources .
figure [ fig : colmag1 ] shows the color - magnitude diagram for the entire sample divided into the different classifications .
the selection criterion for red sequence galaxies is similar to that of @xcite , although we adopt a slightly redder cut to limit contamination and do not employ any optical or infrared apparent magnitude color cuts .
of the 914 red galaxies , 89 ( @xmath210% ) have 24 @xmath1 flux densities , [email protected] ( [ 24@xmath1]@xmath4416.1 mag ) .
we refer to this population as `` 24@xmath1 excess '' red galaxies ( red galaxies with [email protected] mjy are referred to as `` 24@xmath1 faint '' red galaxies ) .
all of these have irac 3.6 @xmath1 detections and we expect all 24 @xmath1 matches to be that of a true counterpart ( we expect only 0.07% of our galaxies to have 24 @xmath1 matches within a radius of 2of the 24@xmath1 position by coincidence ) .
we note that our use of [email protected] to define `` 24@xmath1 excess '' red galaxies is a somewhat arbitrary limit determined by the depth of our mips 24@xmath1 data
. it does however correspond to infrared luminosities of l@xmath46 at @xmath10 = 0.15 ( 0.3 ) that are orders of magnitude higher than we would expect from an old stellar population ( see section [ sec : intro ] ) .
in this section , we use infrared colors , optical line diagnostics , optical morphology , and mean multi - wavelength spectral energy distributions to investigate the nature of the 24@xmath1 excess red galaxies .
only 8/914 ( @xmath441% ) of the red galaxies have x - ray detections in the xbotes chandra survey ( @xcite ; @xcite ) .
the x - ray survey limit corresponds to an x - ray luminosity of 5@xmath47 ( 2 @xmath48 ) ergs s@xmath49 at @xmath10=0.15 ( 0.3 ) .
the small number of detections suggests that any x - ray emission from agn is generally either weak or obscured and so we do not discuss the x - ray emission further . the irac color - color diagram
( e.g. , @xcite ; @xcite ; @xcite ) has been shown to be a reliable way of identifying high redshift agn .
figure [ fig : irac ] shows the irac color - color diagram for the ages galaxies at [email protected] .
the vast majority of sources have [ 3.6]-[4.5 ] colors consistent with them being low redshift galaxies ( e.g. , @xcite ) .
the red and blue galaxies show a striking bimodality in [ 5.8]-[8.0 ] color ( this is also demonstrated in @xcite ) .
the blue galaxies have redder [ 5.8]-[8.0 ] colors which is likely due to the strong 6.2 and 7.7 @xmath51 pah emission features falling within the observed 8 @xmath1 band - pass at these redshifts ( e.g. , @xcite ) .
the red galaxies are likely to have bluer [ 5.8]-[8.0 ] colors because of the weakness or lack of pah emission features and/or the 1.6@xmath1 stellar photospheric feature ( the `` stellar bump '' ) that may still dominate the infrared emission in less active sources with large old stellar populations .
we overplot the color evolution of m82 and ngc 4429 using the templates of @xcite .
m82 is a starburst galaxy with strong pah emission features .
ngc 4429 is an inactive massive s0/sa galaxy in the virgo cluster .
the tracks confirm that actively star - forming and non - star - forming galaxies lie in different regions of the irac color - color diagram at these redshifts .
the 24@xmath1 excess red galaxies lie in the region of the color - color diagram between that of the blue star - forming galaxies and the general red galaxy population ( and in a similar region to that of the green valley sources ) .
there is no correlation of [ 5.8]-[8.0 ] color with f@xmath52 .
the infrared brightest sources fall in both the non - pah and pah regions . to quantify the distribution of sources in figure [ fig : irac ] , we divide the color - color space into 3 regions : the `` agn '' wedge region expected to be populated by powerful agn as defined by @xcite , and the `` pah '' and `` non - pah '' regions .
the boundary between these two populations at an infrared color of [ 5.8]-[8.0 ] = 1.13 mag is defined empirically so that only @xmath25% of the blue galaxies are defined as non - pah sources and @xmath25% of the non-24@xmath1 excess red galaxies are defined as pah sources .
to be classified , we require sources to have a detection in at least one irac band and to appear unambiguously in one region of the irac color - color diagram once limits are taken into account ( otherwise , it is classified as `` unknown '' )
. figure [ fig : hist ] shows the fraction of red , green , blue , and infrared - excess red galaxies classified as powerful agn , pah , or non - pah from figure [ fig : irac ] .
approximately 64% of the 24@xmath1 excess red galaxies are classified as pah sources ( compared to only 5% of the 24@xmath1 faint red galaxies faint red galaxies are unclassifiable .
however , this can not account for the large difference , even in the unlikely event that all of the remaining 28% were classified as pah sources . ] ) .
this result suggests that the infrared emission is largely due to star formation . in figure
[ fig : bpt ] we show an optical line diagnostic diagram ( @xcite ; @xcite ; @xcite ; @xcite ) for all of the ages galaxies .
we consider all 1996 sources with h@xmath36 and [ nii]@xmath536584 emission lines with a signal to noise ratio of s / n@xmath543 ( 51% of the total sample ) .
if a source has an optical spectrum with s / n@xmath543 for all four of the h@xmath36 , h@xmath37 , [ oiii]@xmath535007 , and [ nii]@xmath536584 emission lines , it is classified as an agn or star - forming galaxy using the empirically derived classification line of @xcite .
we also include sources with a signal to noise of s / n@xmath543 on the h@xmath36 and [ nii ] emission lines but with s / n@xmath443 on either of the h@xmath37 and [ oiii ] emission lines .
we classify these sources as an agn if log([nii]/h@xmath55 -0.3 and a star - forming galaxy if log([nii]/h@xmath56 -0.3 ( see @xcite for a similar classification criterion ) .
we classify sources with s / n@xmath543 on the h@xmath36 and s / n@xmath443 on the [ nii ] emission lines and log([nii]@xmath57/h@xmath36 ) @xmath58 -0.3 ( where [ nii]@xmath59 is the 1@xmath31 upper limit on the [ nii ] emission line measurement ) as star - forming galaxies .
any source with s / n@xmath60 on both the h@xmath36 and h@xmath37 emission lines and fwhm@xmath54500 km @xmath61 is classified as a ( broad - line ) agn .
if a source does not meet any of these criteria , it is classified as `` unknown '' .
this results in a completeness rate of 70/89 ( 79% ) , 241/825 ( 29% ) , 377/720 ( 52% ) , and 1942/2255 ( 86% ) for 24@xmath1 excess red galaxies , 24@xmath1 faint red galaxies , green galaxies , and blue galaxies respectively at [email protected] .
blue galaxies tend to have [ oiii]/h@xmath37 versus [ nii]/h@xmath36 line ratios expected for star - forming galaxies whereas red galaxies lie in a much larger region , spanning all classifications .
the 24@xmath1 excess red galaxies fall in a region similar to that of the green valley galaxies .
figure [ fig : bpt_hist ] shows the fraction of 24@xmath1 excess red galaxies , 24@xmath1 faint red galaxies , green galaxies , and blue galaxies classified as star - forming , agn , or unknown ( and with the sparse sampling weights taken into account ) .
galaxies with broad optical lines are included in the agn classification .
57/89 ( 64% ) and 13/89 ( 15% ) of the 24@xmath1 excess red galaxies are classified as agn and star - forming sources respectively ( with 19/89 ( 21% ) unclassifiable ) .
the 24@xmath1 excess red galaxies with infrared colors consistent with pah emission features are not confined to the star - forming region but are distributed throughout the optical line diagnostic diagram .
this result suggests that a large proportion of 24@xmath1 excess red galaxies exhibit characteristics of both agn and star - forming galaxies .
figure [ fig : sed ] shows the mean multi - wavelength spectral energy distribution ( sed ) of the 24@xmath1 excess red galaxies compared to that of the 24@xmath1 faint red galaxies . overplotted
are the sed templates of @xcite for elliptical , spiral , irregular galaxies and agn extended into the uv and infrared ( these modifications will be presented by assef et al . in preparation ) .
the far and near - uv points are obtained from a galaxy evolution explorer ( galex ; @xcite ) survey in the ndwfs botes field .
the mean seds are essentially indistinguishable except in the infrared ( by definition ) and the uv .
the excess of uv emission in the 24@xmath1 excess red galaxies is likely due to the same processes that are giving rise to the 24@xmath1 excess .
the mean sed of the 24@xmath1 faint red galaxies is well fit by an elliptical galaxy template ( the slight excess in the near - uv and at 24@xmath1 can be accounted for by the fact that this class still includes some sources with a modest 24@xmath1 excess which do not satisfy the requirement of [email protected] that they would need to be classified as a 24@xmath1 excess red galaxy ) .
the mean uv to 24@xmath1 flux density ratio of the 24@xmath1 excess red galaxies is a factor of 10(4 ) smaller than that of the irregular galaxy ( agn ) template and twice as high as that of the sbc spiral galaxy template .
this suggests that they contain a larger amount of dust and/or host an older stellar population than typical irregular galaxies .
if the 24 @xmath1 and uv excesses are due largely to agn , then the uv emission must be suppressed by dust .
the amount of dust is not well constrained : an extinction of e(b - v ) of a few tenths would be sufficient to suppress the uv light .
figure [ fig : colmag ] shows the distribution of the 24@xmath1 excess red galaxies in color - magnitude space compared to that of the general red galaxy population .
a kolmogorov - smirnov test shows that the distribution of the 24@xmath1 excess red galaxies ( i.e. , all colored points in figure [ fig : colmag ] ) is not significantly different from that of the overall red galaxy population . however , there are hints of differences in the distribution of 24@xmath1 excess red galaxies with different classifications from the optical line diagnostic diagram : the star - forming generally have bluer @xmath63 colors than the agn sources .
we use the ndwfs @xmath29-band images to investigate the optical morphologies of the 24@xmath1 excess red galaxies .
we find that for 42% of the 24@xmath1 excess red galaxies , the optical images exhibit an elongated structure ( the remaining sources tend to resemble typical bulge - dominated galaxies ) .
the large axial ratios ( @xmath2 5 - 10:1 ) are indicative of an edge - on spiral galaxy ( hereafter referred to as a `` spiral morphology '' ) .
the fraction of 24@xmath1 excess red galaxies with a spiral morphology is much larger than the 11% of 24@xmath1 faint red galaxies with such morphologies and suggests that some fraction of the 24@xmath1 excess red galaxies are reddened edge - on disk galaxies . of the 24@xmath1 excess red galaxies within the pah region of the infrared color - color diagram , 50% have spiral morphologies ( compared to 32% of the `` non - pah '' sources ) .
we also compare the local environments of the 24@xmath1 excess red galaxies to that of the 24@xmath1 faint red galaxies .
the 24@xmath1 excess galaxies show no evidence for a larger fraction of sources with either a disturbed morphology or close companion compared to that of the 24@xmath1 faint red galaxies .
we find that a small but significant minority ( 10% ) of red galaxies have an 24@xmath1 excess .
we have used the irac colors of 24@xmath51 excess red galaxies to show that a large fraction are likely to exhibit pah emission features , indicative of star - formation .
although the infrared emission of elliptical galaxies is usually attributed to agn or stellar mass loss from evolved stars ( e.g. , @xcite ; @xcite ) , pah emission features have recently been detected in nearby elliptical galaxies @xcite .
@xcite find that the 6.2 , 7.7 , and 8.6 @xmath1 emission features are unusually weak in comparison to pah emission features at longer wavelengths .
@xcite show that this peculiar pah emission spectrum with markedly diminished 5 - 8 @xmath1 features arises among the sample solely in systems with relatively hard radiation fields harboring low - luminosity agn . given this effect ,
it is perhaps even more surprising that we find such a large fraction of 24@xmath64 excess red galaxies in the `` pah '' region of the irac color - color diagram .
there appears to be a discrepancy between the optical and infrared diagnostics of the 24@xmath1 excess red galaxies : the optical line diagnostics suggest that agn dominate whereas the infrared colors suggest that star formation dominates .
how do we reconcile these differences ?
this could be , at least in part , an aperture effect .
the optical spectra are obtained using a multi - fiber spectrograph and the fibers subtend 1.5on the sky .
this corresponds to an angular diameter of @xmath325.4 kpc at @xmath10=0.225 .
the irac magnitudes were obtained from 6diameter apertures , corresponding to angular diameter of @xmath3222 kpc at @xmath10=0.225 .
the optical diagnostics are therefore likely to probe only the region of the galaxy bulge , whereas the infrared diagnostics consider infrared light from the entire galaxy . probing only the central regions may result in a larger contribution by an agn , if present .
the optical line diagnostics are also very sensitive to agn activity . in particular ,
@xcite have shown that 80% of optically classified agn have compact radio cores , demonstrating that optical spectroscopy identifies agn with a high rate of success , despite potential problems with dust obscuration .
presumably , in many cases , the extended narrow - line emission is detected even when the nucleus is heavily dust obscured and the infrared colors are dominated by pah emission from circum - nuclear star formation .
it is also likely that different processes may dominate at different wavelengths .
this study provides a caution against using diagnostics in one wavelength to determine what dominates the bolometric luminosity of the source . what is the nature of the 24@xmath1 excess red galaxies ?
we explore three possibilities : + ( 1 ) dusty star - forming galaxies with high extinctions . the 24@xmath1 excess red galaxies could be a population of dusty star - forming galaxies with enough extinction to give them redder @xmath65-@xmath66 colors ( e.g. , @xcite ; @xcite ) . in section [ sec : morph ]
, we show that 42% of the 24@xmath1 excess red galaxies have optical morphologies indicative of edge - on spiral galaxies.this fraction is much larger than the 11% of 24@xmath1 faint red galaxies with such morphologies ( and the @xmath215% of the red sequence population estimated to be dusty galaxies in the hst gems survey at @xmath670.7 @xcite ) .
we compute the h@xmath36/h@xmath37 balmer decrements for all 24@xmath1 excess red galaxies with a signal to noise , s / n@xmath543 on the h@xmath36 and h@xmath37 lines ( 48/89=54% of the total sample ) . of the measurable sources
, we find that @xmath3260% have optical extinctions that are sufficiently large that their intrinsic optical colors are consistent with them being blue or green valley galaxies ( compared to only 13% of the 24@xmath1 faint red galaxies with measurable lines ) .
similar results were reported by @xcite , who find that @xmath250% of their infrared bright red galaxies classified as star - forming galaxies have intrinsic optical colors consistent with them being blue sequence galaxies . however , the large bulge components that they find in their sources suggest that these are not pure disc - like galaxies with high levels of extinction . of the 24@xmath1 excess red galaxies within the pah region of the infrared color - color diagram , 50% have spiral morphologies ( compared to 32% of the `` non - pah '' sources ) .
this result suggests that dusty star - forming galaxies contribute to some , but not all , of the 24@xmath1 excess red galaxy population .
\(2 ) quenched galaxies that have recently joined the red sequence from the blue cloud .
the 24@xmath1 excess red galaxies could represent a transition population of sources whose star - formation has been recently quenched ( in @xmath441 gyr ; @xcite ; @xcite ) .
the population exhibits characteristics that would be consistent with such a scenario .
one could envisage a situation in which the 24@xmath1 excess red galaxies that show star - forming characteristics in their optical line diagnostics are the sources that are in the earlier stages of being quenched .
as the sources become quenched further , they become redder in their @xmath65-@xmath66 colors , and their optical line diagnostics become dominated by agn . in figure
[ fig : colmag ] , we show the color magnitude diagram for the 24@xmath1 excess red galaxies split into star - forming sources and agn as classified from their optical line diagnostics .
the star - forming 24@xmath1 excess red galaxies appear to be slightly bluer on average than the sources classified as agn .
+ ( 3 ) merging red galaxies .
the 24@xmath1 excess red galaxy population could be massive red galaxies that are undergoing major or minor mergers and exhibiting obscured star - formation and agn activity .
while the optical line diagnostics suggest that low level agn activity appears to be common in both 24@xmath1 excess and 24@xmath1 faint red galaxies , the 24@xmath1 excess red galaxies may indicate a recent galaxy merger or interaction that triggers obscured star - formation activity that is only detectable in the infrared ( and marginally in the uv which is more sensitive to star - formation activity than the optical ) .
this may lower the need for `` dry '' mergers in massive red galaxies .
the distribution of stellar masses ( estimated from the optical photometry ) for the 24@xmath1 excess red galaxies ( @xmath68 ) shows no significant difference from that of the 24@xmath1 faint red galaxies .
in addition , the infrared luminosity is poorly correlated with the stellar mass estimated from the optical photometry ( j. moustakas et al .
, in preparation ) , consistent with what we would expect if there are brief episodes of activity occurring within the same parent population .
these results are more consistent with the 24@xmath1 excess population being dominated by merging red galaxies rather than quenched galaxies that have recently joined the red sequence . in the latter case , the `` downsizing '' of galaxies ( in which galaxies of decreasing mass are progressively joining the red sequence with time ) would imply that the masses of 24@xmath1 excess red galaxies should be less than that of 24@xmath1 faint red galaxies ( that joined the red sequence at a higher redshift ) . on the other hand , we show in section [ sec : morph ] that the 24@xmath1 excess galaxies show no evidence for a larger fraction of sources with either a disturbed morphology or close companion than that of the 24@xmath1 faint red galaxies .
+ it appears that the low redshift 24@xmath1 excess red galaxy population comprises a mixture of different types of sources .
roughly 40% are likely to be edge - on disk galaxies with high extinction .
the rest are likely to be red galaxies whose 24@xmath1 emission comes from both obscured agn and star formation activity . they may be a transition population of sources that have only recently left the blue cloud to join the red sequence and in which the residual star - formation is being quenched by agn activity that becomes increasingly dominant in the optical band as the sources become redder . alternatively , they could be massive red galaxies undergoing a burst of obscured star - formation activity ( perhaps due to a merger ) .
further work needs to be done to determine which of these mechanisms dominates and how this changes with redshift .
by analyzing the infrared colors and optical line diagnostics of [email protected] 24@xmath1 excess red galaxies and comparing these to that of blue , green valley , and 24@xmath1 faint red galaxies within the ages survey , we find the following : * of the total red galaxy population , 10% have 24 @xmath70 flux densities , [email protected] mjy .
these correspond to infrared luminosities that are larger than we would expect from the infrared emission of local elliptical galaxies . * using their [ 5.8]-[8.0 ] irac colors , we show that 64% of the 24@xmath1 excess red galaxies are likely to have strong pah emission features in their infrared spectra ( compared to only 5% of the 24@xmath1 faint red galaxies ) .
this result suggests that in a large fraction of the sources , the infrared emission is dominated by star - formation processes . * a much larger fraction of 24@xmath1 excess red galaxies are classified as star - forming using infrared diagnostics ( @xmath264% ) than optical line diagnostics ( @xmath215% ) .
this result suggests that in many sources , both agn and star - formation activity are occurring simultaneously ( but that in the optical , the agn dominates the signal whereas in the infrared , the pah emission features dominate . ) * although the optical to near - ir seds of the 24@xmath1 excess and 24@xmath1 quiet galaxies are very similar , the 24@xmath1 excess red galaxies appear to also have a small uv excess over that of the 24@xmath1 quiet galaxies .
the small uv to 24@xmath1 ratio suggests a large amount of dust and/or relatively old stellar population compared to that of irregular galaxies . *
approximately 40% of 24@xmath1 excess red galaxies at low redshifts are likely to be dusty edge - on spiral galaxies , whose high optical extinctions make them red .
the rest are likely to be red galaxies whose 24@xmath1 emission comes from both obscured agn and star formation activity .
this may be triggered by minor mergers of red galaxies or , in some cases , may be the residual activity as blue galaxies are quenched to form red sequence galaxies .
we thank our colleagues on the ndwfs , mips , ages , and irac teams .
kb is supported by the giacconi fellowship at stsci .
jm acknowledges funding support from nasa-06-galex06 - 0030 and spitzer g05- ar-50443 .
this research is partially supported by the national optical astronomy observatory which is operated by the association of universities for research in astronomy , inc .
( aura ) under a cooperative agreement with the national science foundation .
this work is based on observations made with the _ spitzer space telescope _ , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa .
the _ spitzer / mips _ survey of the botes region was obtained using gto time provided by the _ spitzer _ infrared spectrograph team ( james houck , p.i . ) and by m. rieke .
ages is a collaboration between scientists at the harvard - smithsonian center for astrophysics , steward observatory , the national optical astronomical observatory , the ohio state university , and the jet propulsion laboratory.we thank the anonymous referee for his / her useful comments . , r. j. , kochanek , c. s. , brodwin , m. , brown , m. j. i. , caldwell , n. , cool , r. j. , eisenhardt , p. , eisenstein , d. , gonzalez , a. h. , jannuzi , b. t. , jones , c. , mckenzie , e. , murray , s. s. , & stern , d. 2008 , , 676 , 286 , e. f. , mcintosh , d. h. , barden , m. , wolf , c. , caldwell , j. a. r. , rix , h .- w . , beckwith , s. v. w. , borch , a. , hussler , b. , jahnke , k. , jogee , s. , meisenheimer , k. , peng , c. , sanchez , s. f. , somerville , r. s. , & wisotzki , l. 2004 , , 600 , l11 , k. , brown , m. j. i. , dey , a. , jannuzi , b. t. , kochanek , c. s. , kenter , a. t. , fabricant , d. , fazio , g. g. , forman , w. r. , green , p. j. , jones , c. j. , mcnamara , b. r. , murray , s. s. , najita , j. r. , rieke , m. , shields , j. c. , & vikhlinin , a. 2006 , , 641 , 140 , k. , dey , a. , brown , m. j. i. , watson , c. r. , jannuzi , b. t. , najita , j. r. , kochanek , c. s. , shields , j. c. , fazio , g. g. , forman , w. r. , green , p. j. , jones , c. j. , kenter , a. t. , mcnamara , b. r. , murray , s. s. , rieke , m. , & vikhlinin , a. 2005 , , 626 , 723 , p. , pozzi , f. , oliver , s. , polletta , m. , afonso - luis , a. , farrah , d. , hatziminaoglou , e. , rodighiero , g. , berta , s. , waddington , i. , lonsdale , c. , rowan - robinson , m. , shupe , d. l. , evans , t. , fang , f. , smith , h. e. , & surace , j. 2006 , , 371 , 1113 , p.
r. , stern , d. , brodwin , m. , fazio , g. g. , rieke , g. h. , rieke , m. j. , werner , m. w. , wright , e. l. , allen , l. e. , arendt , r. g. , ashby , m. l. n. , barmby , p. , forrest , w. j. , hora , j. l. , huang , j .- s .
, huchra , j. , pahre , m. a. , pipher , j. l. , reach , w. t. , smith , h. a. , stauffer , j. r. , wang , z. , willner , s. p. , brown , m. j. i. , dey , a. , jannuzi , b. t. , & tiede , g. p. 2004
, , 154 , 48 , s. m. , willmer , c. n. a. , wolf , c. , koo , d. c. , weiner , b. j. , newman , j. a. , i m , m. , coil , a. l. , conroy , c. , cooper , m. c. , davis , m. , finkbeiner , d. p. , gerke , b. f. , gebhardt , k. , groth , e. j. , guhathakurta , p. , harker , j. , kaiser , n. , kassin , s. , kleinheinrich , m. , konidaris , n. p. , kron , r. g. , lin , l. , luppino , g. , madgwick , d. s. , meisenheimer , k. , noeske , k. g. , phillips , a. c. , sarajedini , v. l. , schiavon , r. p. , simard , l. , szalay , a. s. , vogt , n. p. , & yan , r. 2007 , , 665 , 265 , d. , fata , r. , roll , j. , hertz , e. , caldwell , n. , gauron , t. , geary , j. , mcleod , b. , szentgyorgyi , a. , zajac , j. , kurtz , m. , barberis , j. , bergner , h. , brown , w. , conroy , m. , eng , r. , geller , m. , goddard , r. , honsa , m. , mueller , m. , mink , d. , ordway , m. , tokarz , s. , woods , d. , wyatt , w. , epps , h. , & dellantonio , i. 2005 , , 117 , 1411 , d. g. , hertz , e. n. , szentgyorgyi , a. h. , fata , r. g. , roll , j. b. , & zajac , j. m. 1998 , in presented at the society of photo - optical instrumentation engineers ( spie ) conference , vol . 3355 , proc .
3355 , p. 285
- 296 , optical astronomical instrumentation , sandro dodorico ; ed . , ed .
s. dodorico , 285296 , g. g. , hora , j. l. , allen , l. e. , ashby , m. l. n. , barmby , p. , deutsch , l. k. , huang , j .- s . , kleiner , s. , marengo , m. , megeath , s. t. , melnick , g. j. , pahre , m. a. , patten , b. m. , polizotti , j. , smith , h. a. , taylor , r. s. , wang , z. , willner , s. p. , hoffmann , w. f. , pipher , j. l. , forrest , w. j. , mcmurty , c. w. , mccreight , c. r. , mckelvey , m. e. , mcmurray , r. e. , koch , d. g. , moseley , s. h. , arendt , r. g. , mentzell , j. e. , marx , c. t. , losch , p. , mayman , p. , eichhorn , w. , krebs , d. , jhabvala , m. , gezari , d. y. , fixsen , d. j. , flores , j. , shakoorzadeh , k. , jungo , r. , hakun , c. , workman , l. , karpati , g. , kichak , r. , whitley , r. , mann , s. , tollestrup , e. v. , eisenhardt , p. , stern , d. , gorjian , v. , bhattacharya , b. , carey , s. , nelson , b. o. , glaccum , w. j. , lacy , m. , lowrance , p. j. , laine , s. , reach , w. t. , stauffer , j. a. , surace , j. a. , wilson , g. , wright , e. l. , hoffman , a. , domingo , g. , & cohen , m. 2004 , , 154 , 10 , v. , brodwin , m. , kochanek , c. s. , murray , s. , stern , d. , brand , k. , eisenhardt , p. r. , ashby , m. l. n. , barmby , p. , brown , m. j. i. , dey , a. , forman , w. , jannuzi , b. t. , jones , c. , kenter , a. t. , pahre , m. a. , shields , j. c. , werner , m. w. , & willner , s. p. 2008
, , 679 , 1040 , b. t. , & dey , a. 1999 , in astronomical society of the pacific conference series , vol .
191 , photometric redshifts and the detection of high redshift galaxies , ed .
r. weymann , l. storrie - lombardi , m. sawicki , & r. brunner , 111+ , g. , heckman , t. m. , tremonti , c. , brinchmann , j. , charlot , s. , white , s. d. m. , ridgway , s. e. , brinkmann , j. , fukugita , m. , hall , p. b. , ivezi , . , richards , g. t. , & schneider , d. p. 2003
, , 346 , 1055 , a. , murray , s. s. , forman , w. r. , jones , c. , green , p. , kochanek , c. s. , vikhlinin , a. , fabricant , d. , fazio , g. , brand , k. , brown , m. j. i. , dey , a. , jannuzi , b. t. , najita , j. , mcnamara , b. , shields , j. , & rieke , m. 2005 , , 161 , 9 , m. , storrie - lombardi , l. j. , sajina , a. , appleton , p. n. , armus , l. , chapman , s. c. , choi , p. i. , fadda , d. , fang , f. , frayer , d. t. , heinrichsen , i. , helou , g. , i m , m. , marleau , f. r. , masci , f. , shupe , d. l. , soifer , b. t. , surace , j. , teplitz , h. i. , wilson , g. , & yan , l. 2004 , , 154 , 166 , c. j. , smith , h. e. , rowan - robinson , m. , surace , j. , shupe , d. , xu , c. , oliver , s. , padgett , d. , fang , f. , conrow , t. , franceschini , a. , gautier , n. , griffin , m. , hacking , p. , masci , f. , morrison , g. , olinger , j. , owen , f. , prez - fournon , i. , pierre , m. , puetter , r. , stacey , g. , castro , s. , polletta , m. d. c. , farrah , d. , jarrett , t. , frayer , d. , siana , b. , babbedge , t. , dye , s. , fox , m. , gonzalez - solares , e. , salaman , m. , berta , s. , condon , j. j. , dole , h. , & serjeant , s. 2003 , , 115 , 897 , d. c. , fanson , j. , schiminovich , d. , morrissey , p. , friedman , p. g. , barlow , t. a. , conrow , t. , grange , r. , jelinsky , p. n. , milliard , b. , siegmund , o. h. w. , bianchi , l. , byun , y .- i . , donas , j. , forster , k. , heckman , t. m. , lee , y .- w . , madore , b. f. , malina , r. f. , neff , s. g. , rich , r. m. , small , t. , surber , f. , szalay , a. s. , welsh , b. , & wyder , t. k. 2005 , , 619 , l1 , g. h. , young , e. t. , engelbracht , c. w. , kelly , d. m. , low , f. j. , haller , e. e. , beeman , j. w. , gordon , k. d. , stansberry , j. a. , misselt , k. a. , cadien , j. , morrison , j. e. , rivlis , g. , latter , w. b. , noriega - crespo , a. , padgett , d. l. , stapelfeldt , k. r. , hines , d. c. , egami , e. , muzerolle , j. , alonso - herrero , a. , blaylock , m. , dole , h. , hinz , j. l. , le floch , e. , papovich , c. , prez - gonzlez , p. g. , smith , p. s. , su , k. y. l. , bennett , l. , frayer , d. t. , henderson , d. , lu , n. , masci , f. , pesenson , m. , rebull , l. , rho , j. , keene , j. , stolovy , s. , wachter , s. , wheaton , w. , werner , m. w. , & richards , p. l. 2004 , , 154 , 25 , j. d. t. , draine , b. t. , dale , d. a. , moustakas , j. , kennicutt , jr .
, r. c. , helou , g. , armus , l. , roussel , h. , sheth , k. , bendo , g. j. , buckalew , b. a. , calzetti , d. , engelbracht , c. w. , gordon , k. d. , hollenbach , d. j. , li , a. , malhotra , s. , murphy , e. j. , & walter , f. 2007 , , 656 , 770 , d. , eisenhardt , p. , gorjian , v. , kochanek , c. s. , caldwell , n. , eisenstein , d. , brodwin , m. , brown , m. j. i. , cool , r. , dey , a. , green , p. , jannuzi , b. t. , murray , s. s. , pahre , m. a. , & willner , s. p. 2005
, , 631 , 163 , i. , ivezi , . , knapp , g. r. , narayanan , v. k. , strauss , m. a. , gunn , j. e. , lupton , r. h. , schlegel , d. , bahcall , n. a. , brinkmann , j. , brunner , r. j. , budavri , t. , csabai , i. , castander , f. j. , doi , m. , fukugita , m. , gyry , z. , hamabe , m. , hennessy , g. , ichikawa , t. , kunszt , p. z. , lamb , d. q. , mckay , t. a. , okamura , s. , racusin , j. , sekiguchi , m. , schneider , d. p. , shimasaku , k. , & york , d. 2001 , , 122 , 1861 , m. w. , roellig , t. l. , low , f. j. , rieke , g. h. , rieke , m. , hoffmann , w. f. , young , e. , houck , j. r. , brandl , b. , fazio , g. g. , hora , j. l. , gehrz , r. d. , helou , g. , soifer , b. t. , stauffer , j. , keene , j. , eisenhardt , p. , gallagher , d. , gautier , t. n. , irace , w. , lawrence , c. r. , simmons , l. , van cleve , j. e. , jura , m. , wright , e. l. , & cruikshank , d. p. 2004
, , 154 , 1 color - magnitude diagram for all @xmath71 - 19.4 mag ages galaxies at @xmath72 .
the sources are divided into red sequence galaxies ( filled red circles ) , green valley galaxies ( empty green circles ) , blue cloud galaxies ( greyscale ) . also shown
is a histogram of the distribution of u - v colors of the sources .
the dotted lines show the selection criteria at @xmath73=-20 mag and @xmath10=0.225.,height=302 ] irac color - color diagram for all ages galaxies .
the greyscale shows the distribution of blue galaxies .
the red circles , green squares , and yellow stars denote red , green valley , and 24@xmath1 excess red galaxies respectively .
the larger yellow stars correspond to 24@xmath1 excess red galaxies with larger 24@xmath1 flux densities . for clarity , for the blue , green , and red galaxies , only the sources with detections in all 4 irac bands are shown . for the 24@xmath1 excess red galaxies , where an irac flux is below the detection limit
, the limits on the irac colors are shown .
over - plotted are the non - evolving [email protected] color tracks for m82 ( a starburst galaxy ; dotted black line ) and ngc 4429 ( a massive inactive s0/sa galaxy ; solid black line ) . in both cases ,
the redshift increases upwards and to the left .
the diagram is divided into three sections by the dashed black line : the powerful agn `` wedge '' as defined by @xcite , and the `` pah '' and `` non - pah '' regions defined empirically from the distribution of the red and blue galaxies.,height=302 ] classification of galaxies based on the irac color - color diagram ( figure [ fig : irac ] ) .
the histogram shows the fraction of 24@xmath1 excess red galaxies , 24@xmath1 faint red galaxies , green galaxies , and blue galaxies within the pah ( blue slanted lines ) , non - pah ( solid red shading ) , and powerful agn ( green vertical lines ) regions or that are unclassifiable ( `` unknown '' ; white area ) .
the fractions have been corrected for the ages incompleteness .
the fractions of 24@xmath1 excess red galaxies , 24@xmath1 faint red galaxies , green galaxies , and blue galaxies within the agn ( unclassifiable ) regions are 3% ( 2% ) , 2% ( 29% ) , 4% ( 28% ) , and 5% ( 18% ) respectively . , height=302 ] the log([oiii]@xmath535007/h@xmath37 ) vs. log([nii]@xmath536584/h@xmath36 ) line diagnostic diagram for galaxies with measurable line fluxes .
the symbols are as defined in figure [ fig : irac ] .
the empirically derived line of @xcite ( solid line ) used to distinguish between star - forming galaxies and narrow - line agn is shown as a solid line . the @xcite extreme starburst classification line ( dashed line ) is also shown for reference .
the line at log([nii]/h@xmath36=-0.3 used to classify sources with no suitable h@xmath37 and/or [ nii ] measurement is also shown ( dotted line ) .
for the 24@xmath1 excess galaxies , arrows show the limits on the line ratios in cases where lines are not detected.,height=302 ] classification of galaxies based on their optical line diagnostics ( see figure [ fig : bpt ] ) .
the histogram shows the fraction of 24@xmath1 excess red galaxies , 24@xmath1 faint red galaxies , green galaxies , and blue galaxies within the star - forming ( `` sf'';blue slanted lines ) and agn ( `` agn'';solid red shading ) regions of the log([oiii]/h@xmath37 ) vs. log([nii]/h@xmath36 line diagnostic diagram and sources whose optical spectra are not good enough to classify the sources ( `` unknown '' ; white area ) .
the fractions have been corrected for the ages incompleteness.,height=302 ] the average multi - wavelength spectral energy distribution of the 24@xmath1 excess red galaxies ( red triangles ) and 24@xmath1 faint red galaxies ( black squares ) . also shown
are the errors on the mean flux density .
overplotted are sed templates from @xcite .
the solid black line represents an elliptical galaxy template normalized to the peak of the emission from the 24@xmath1 faint red galaxies .
the dotted red line ( dashed blue line ; long - dashed green line ) shows the sum of the sbc spiral template ( irr template ; agn template ) and the elliptical galaxy template that simultaneously match the 24@xmath1 emission of the 24@xmath1 excess red galaxies and the peak of the stellar emission.,height=340 ] color - magnitude diagram showing rest - frame @xmath65-@xmath66 color vs absolute @xmath66-band magnitude ( vega ) .
the greyscale shows the distribution of all red sequence galaxies .
all symbols represent 24@xmath1 excess red galaxies . filled red stars and filled blue circles denote 24@xmath1 excess red galaxies classified as agn and star - forming galaxies respectively from the optical line diagnostic diagram ( @xcite ; @xcite ; @xcite ; see figure [ fig : bpt ] ) .
the black dots denote 24@xmath1 excess red galaxies classified as having pah emission from the irac color - color diagram ( see figure [ fig : irac ] ) .
, height=302 ] | observations with the @xmath0 have revealed a population of red sequence galaxies with a significant excess in their 24@xmath1 emission compared to what is expected from an old stellar population .
we identify @xmath2900 red galaxies with @xmath3 from the agn and galaxy evolution survey ( ages ) selected from the noao deep wide - field survey botes field . using @xmath4 mips , we classify 89 ( @xmath210% ) with 24@xmath1 infrared excess ( [email protected] ) .
we determine the prevalence of agn and star - formation activity in all the ages galaxies using optical line diagnostics and mid - ir color - color criteria . using the irac color - color diagram from the irac shallow survey
, we find that 64% of the 24@xmath1 excess red galaxies are likely to have strong pah emission features in the 8@xmath1 irac band .
this fraction is significantly larger than the 5% of red galaxies with [email protected] that are estimated to have strong pah emission , suggesting that the infrared emission is largely due to star - formation processes .
only 15% of the 24@xmath1 excess red galaxies have optical line diagnostics characteristic of star - formation ( 64% are classified as agn and 21% are unclassifiable ) .
the difference between the optical and infrared results suggest that both agn and star - formation activity is occurring simultaneously in many of the 24@xmath1 excess red galaxies .
these results should serve as a warning to studies that exclusively use optical line diagnostics to determine the dominant emission mechanism in the infrared and other bands .
we find that @xmath240% of the 24@xmath1 excess red galaxies are edge - on spiral galaxies with high optical extinctions .
the remaining sources are likely to be red galaxies whose 24@xmath1 emission comes from a combination of obscured agn and star - formation activity . |
current models of galaxy formation and assembly can be characterised by two extreme and competing views .
one such view is represented by the early `` monolithic collapse '' model , whereby the spheroidal ( bulge ) components of galaxies form by the early collapse of individual gas clouds ( e.g. larson 1974 ; carlberg 1984 ) . in these early collapse models the star formation induced by the collapse is terminated when the source of gas is either consumed or expelled from the galaxy .
the bulge then ages passively ( albeit perhaps subject to the later addition of a disk ) .
such models are largely motivated by the small scatter in the observed relationships between colour and magnitude ( e.g. baum 1959 ; bower , lucy & ellis 1992a , b ) and velocity dispersion , surface brightness and size ( the fundamental plane ; dressler et al .
1987 ) for early - type galaxies and bulges . to explain the tightness of these relations , early collapse models require that star formation in bulges terminates at early times ( z@xmath52 ; bower , lucy & ellis 1992a , b ) , so that bulge dominated ( e.g. elliptical ) galaxies should be uniformly old .
this picture faces challenges on a number of fronts .
for instance , the differing slopes found for the fundamental plane in various bands ( e.g. bernardi et al .
2003b ) are difficult to explain in the monolithic collapse scenario .
furthermore , spectroscopic studies of the stellar populations in the bulges of galaxies find central ages ranging from @xmath0115 gyr ( e.g. trager et al .
2000 ; proctor & sansom 2002 ; terlevich & forbes 2002 : caldwell , rose & concannon 2003 ; kauffmann et al .
2003a , b ; proctor et al .
2004b ) , indicating evidence for more recent star formation .
however , by far the most important challenge to the early collapse picture is the fundamental differences in galaxy formation and assembly histories that are predicted when the growth of structure via gravitational instability is modelled for a dark matter - dominated universe , now characterised extremely well via precision measurements of the cosmological parameters ( e.g. the @xmath6cdm model ; spergel et al .
2003 ) . such `` hierarchical '' models , which represent the other extreme view , predict galaxy - mass objects to be assembled by the successive merging of lower - mass objects via hierarchical merging ( white & rees 1978 ) , usually over an extended period . another important prediction of hierarchical merging models is that the merger rate of galaxies is a function of environment , i.e. on average , galaxies in the dense environs of clusters merge early , while those in the less dense ` field ' merge later ( kauffmann 1996 ) . in the hierarchical merging scenario
the properties of galaxies are therefore subject to strong influences from their environment .
results from stellar population studies appear to confirm the prediction of hierarchical merging models in that ` field ' galaxies possess slightly younger central ages than their cluster counterparts ( e.g. thomas et al .
, 2005 ; bernardi et al . , 2006 ; smith et al . ,
2006 ) .
in contrast to early collapse models , hierarchical merging models predict galaxy bulges to possess a range of ages , with low - mass galaxies forming earlier , and on shorter timescales , than high - mass galaxies . at first glance
, the wide range of ages observed in spectroscopic studies might also appear to confirm the hierarchical merging predictions , however the observed trend of age with mass is for lower - mass systems to be _ younger _ than higher - mass systems .
this is inconsistent with hierarchical merging models which predict lower - mass systems to form earlier than high - mass system ( kauffmann 1996 ) .
the observed trend is therefore sometimes referred to as ` anti - hierarchical ' .
the recent de lucia et al .
( 2006 ) analysis of the millennium simulation provides insights important to the development of a coherent picture of galaxy evolution .
the millennium simulation models the evolution of dark matter using an n - body code to which de lucia et al .
couple a semi - analytic model of galaxy formation .
de lucia et al s analysis clearly demonstrates that a strong distinction must be made between the _ mass assembly _ of galaxies and the _ formation _ of their stellar populations .
this can be understood intuitively by the simple realisation that mergers bring old stars into a merger remnant as well as making new stars .
merger remnants may therefore be , and often are , still dominated ( in terms of their mass ) by old stellar populations .
de lucia et al . conclude that while the mass assembly of galaxies is indeed hierarchical , their star formation histories are nevertheless anti - hierarchical . despite this improvement in our theoretical understanding ,
the problem remains how to reconcile the broad range of predicted ( and spectroscopically observed ) galaxy ages with the extremely small scatter in some scaling relations .
recent surveys of large numbers of nearby galaxies have provided statistically significant results and hence further observational constraints on galaxy formation models .
such surveys include the 2df galaxy redshift survey ( 2dfgrs : colless et al .
2001b ) and the sloan digital sky survey ( sdss ; york et al .
2000 ) . from central region spectra and global optical colours
the 2dfgrs probed the cosmic star formation history at relatively low redshift ( baldry et al .
2002 ) and its variation with local environmental density ( lewis et al .
the sdss also obtained optical colours and central spectra for a large number of low redshift galaxies .
this ongoing dataset has been used by several authors to study the star formation history and scaling relations of nearby galaxies .
these studies include bernardi et al .
( 2003b ) who investigated scaling relations for @xmath09,000 galaxies , eisenstein et al .
( 2003 ) who combined spectra of 22,000 massive galaxies , gallazzi et al .
( 2005 ) who probed the star formation history of @xmath040,000 galaxies , cid fernandes et al .
( 2005 ) who conducted a spectral synthesis of @xmath050,000 galaxies , and chang et al .
( 2006 ) who combined spectra and optical colours with 2mass near - infrared colours for @xmath03,000 galaxies .
such large datasets have allowed these authors to investigate trends as a function of mass , environment , galaxy type etc .
already , a galaxy mass - metallicity relation has been detected in the metallicities of both gas ( tremonti et al .
2004 ) and stars ( gallazzi et al . 2005 ) and strong trends with galaxy environment identified ( lewis et al . 2002 ) .
however , the above studies have not explicitly explored the link between the _ central _ stellar populations , which they observe , and the _ global _ photometry of the galaxies in their samples .
it is this issue that this paper specifically addresses . to this end
, we compare the results of stellar population analysis of spectra from the 6df galaxy survey ( 6dfgs ; jones et al .
2004 ) to the global near - infrared photometry of the 2mass extended source catalog ( 2mass xsc ; jarrett et al . 2000a , b ) and optical supercosmos ( hambly , irwin & macgillivray .
2001 ) data for several thousand galaxies .
we explore such issues as the fraction of galaxies stellar populations that can have been formed in recent star formation events and the effects of stellar populations on the slopes of colour - magnitude relations and the varying ` tilt ' of the fundamental plane .
the paper is laid out as follows . in the next section we outline the photometric and spectroscopic data . in section [ spectroscopicanal ]
we detail the spectroscopic data reductions and analysis used to derive stellar population parameters .
section [ results ] presents the results of our analysis , which are discussed and summarised in section [ concs ] .
the 6dfgs is a spectroscopic survey of the entire southern sky more than 10 degrees from the galactic plane .
the survey measured redshifts for more than 124,000 galaxies .
the primary survey targets were selected primarily from the 2mass extended source catalog ( xsc ) .
all galaxies brighter than k@xmath7=12.75 were included in the sample .
secondary samples of 2mass and supercosmos galaxies , complete down to magnitude limits ( h , j , r@xmath8 , b@xmath9 ) = ( 13.00 , 13.75 , 15.60 , 16.75 ) , were also included .
the survey was carried out on the uk schmidt telescope using the six - degree field ( 6df ) multi - fibre spectrograph .
the 6df instrument simultaneously observed up to 150 spectra over a 5.7@xmath10 field .
the science fibres of the instrument were 6.7 arcsec in diameter .
the spectrograph is bench mounted in an enclosure inside the telescope dome , which gives it increased instrumental stability compared to a spectrographs mounted on a telescope ( e.g. 2df and the sdss spectrograph ) .
this increased instrumental stability maintains consistent instrumental resolution , which is important when working with lick indices .
the survey obtained spectra in the visual ( v ) and red ( r ) wavelength ranges providing coverage from @xmath04000 to @xmath08400 .
the spectra have 5 - 6 fwhm resolution in v and 9 - 12 resolution in r ( see jones et al .
2004 for more details ) .
the main science aims of the survey included : to measure the luminosity function of near - infrared selected galaxies as a function of environment and galaxy type ( jones et al .
2006 ) , to measure the clustering of galaxies , to produce detailed maps of the density and peculiar velocity fields of the nearby universe ( jones et al . 2004
; erdogdu et al .
2006 ) and to study the properties of the underlying stellar population of the galaxies such as ages and chemical abundances .
it is this latter issue to which this paper is addressed . the 6dfgs is particularly suited to the aims of the current study because the selection criteria using 2mass near - infrared photometry provides a sample less biased toward bright , young galaxies than surveys based on optical catalogues .
the k band photometry also has the useful property of being relatively insensitive to both metallicity the presence of dust ( unlike optical bands ) , and is therefore an good tracer of mass .
it also provides a useful tool in breaking the age metallicity degeneracy .
finally , 6dfgs spectra are obtained from region 2 to 3 times larger on the sky than those used in the 2dfgrs and sdss surveys . as a result ,
at the mean redshifts of the surveys , 2df , sdss and 6dfgs fibres correspond to 2.6 , 3.9 and 4.6 kpc / h respectively .
the 6dfgs spectra therefore sample a significantly larger fraction of the total galaxy light than either 2df or sdss spectra .
the spectroscopic data presented here are from the 6dfgs first data release ( jones et al .
lick index , velocity dispersion and emission line - strength measurements were made on all galaxies with redshift quality @xmath11 ( the ` reliable ' redshifts ; jones et al .
this gave a total of 39649 galaxies with 3859 of these observed more than once ( 420 galaxies observed more than twice ) .
if a galaxy was observed more than once , only the highest signal - to - noise - ratio spectrum was used for the measurements .
however , repeat observations were used to establish the errors ( see section [ error_measurement ] ) .
only galaxies with median signal - to - noise ratios @xmath512 @xmath12 were deemed suitable for age and metallicity determinations .
such a value lies at the lower limit of signal - to - noise - ratio that can be used to derive ages and metallicities with precision .
the above selection criteria resulted in @xmath07000 galaxies suitable for lick index measurements .
it should be noted that , as no morphological selection was made , the results presented here include galaxies of the full range of early and late types .
+ the k band magnitudes used in this work were taken to be the k@xmath7 values estimated from the 2mass data by jones et al .
we also use the re - calibrated supercosmos b and r band magnitudes ( hambly et al . 2001 ) .
galaxy radii were taken as the 2mass 20th magnitude per arcsec@xmath13 isophotal radius ( r@xmath14 ; jarrett et al .
we also use k band half - light ( effective ) radii ( r@xmath15 ) for a spectroscopically selected early - type galaxy subsample , as supplied by lachlan campbell ( private communication ) .
absolute magnitudes and physical galaxy radii were calculated using distances based on the measured redshifts and a hubble constant of 70 km s@xmath12 mpc@xmath12 ( our median redshift is 0.035 ) .
errors on the distance estimates were assumed to be a combination of recession velocity uncertainty and peculiar motions . in order to be conservative , we assumed the combined uncertainty to be @xmath161000 km s@xmath12 .
the distributions of some key parameters in the final sample are shown in fig .
[ sample ] .
the sample covers of order 5 magnitudes in luminosity ( @xmath17m@xmath18@xmath19 mag and @xmath20m@xmath21@xmath22 mag ) .
this corresponds to a dynamical mass range of more than two orders of magnitude ( from @xmath2310@xmath24 to @xmath2510@xmath26 m@xmath27 ; section [ photo ] ) . from the sample of @xmath07000 galaxies
suitable for age
metallicity determination ( section [ specsample ] ) @xmath06000 also had a full set of 2mass and supercosmos photometry .
7000 galaxies presented here .
the signal - to - noise - ratio per is estimated as the median of the 6dfgs spectra .
apparent magnitudes are from 2mass and supercosmos data .
absolute magnitudes were calculated using redshifts from the 6dfgs spectra and a hubble constant of 70 km s@xmath12mpc@xmath12 .
k band 20th magnitude isophotal radius is from 2mass .
effective radii are also presented for a ( @xmath025% ) subsample of our 6dfgs galaxies . dashed
lines in the radius plots represent the radius of the 6df aperture ( 3.35 arcsec ) .
, width=302 ]
we measured emission line - strengths for h@xmath28 , h@xmath29 , [ nii]@xmath306548 , [ nii]@xmath306584 , [ oiii]@xmath304959 and [ oiii]@xmath305007 from the 6dfgs spectra . in order to measure the h@xmath31 and [ nii ] emission , a de - blending option in the iraf task _ splot _
this option allows the fitting of gaussians to a series of spectral lines and the measurement of the value of the equivalent widths of each .
the strengths of h@xmath31 and the two [ nii ] doublet lines either side of it were measured in this way from continuum divided spectra .
this process can lead to false measurements where noise spikes or the continuum level are fit with gaussian profiles rather than real features . by consideration of the position and fwhm of the gaussian fit to the feature
, one can reasonably discriminate between real emission features and false detections .
if the line centre was more than 7 away from the features expected wavelength ( given the 6dfgs redshift ) , the measurement was discarded .
if the fwhm was well below the instrumental resolution of the 6df instrument ( @xmath23 5 ) the measurement was also discarded .
finally , extremely large fwhm ( @xmath25 12 ) were discarded .
these large fwhm are mostly gaussian fits to features in the continuum , but may also have occasionally included real features of broad - lined active galactic nuclei ( agn ) . the equivalent widths of the two [ oiii ] emission lines at 4959 and 5007 were measured using the procedure and band definitions of gonzalez ( 1993 ) .
h@xmath29 emission , on the other hand , was estimated for each galaxy as the difference between the _ observed _
6dfgs h@xmath29 index value and the value predicted by our best - fit model solutions ( see section [ fitting ] ) . within this work
the sample is sometimes subdivided into ` passive ' ( no emission ) and ` emission ' galaxies .
passive galaxies are defined as those in which no emission lines are detected within the 6dfgs wavelength range with greater than approximately 1-@xmath32 significance .
emission line galaxies are therefore defined as those in which at least one emission line is detected at greater than one sigma confidence . for h@xmath28 this criterion corresponds to emission - line equivalent widths of @xmath01.5 . using this definition
the sample contains 4640 passive galaxies and 2270 emission galaxies .
unfortunately , a full analysis of morphological types is not available for the sample .
however , visual inspection of a sub - sample of passive galaxies suggests that @xmath080% are early types , while amongst the emission galaxies a similar fraction are late types .
the method used to measure lick indices from appropriately broadened spectra and to correct the indices for the broadening effect of a galaxy s velocity dispersion was based on that used by stephen moore ( 2001 ) measurements of the velocity dispersion in the 6dfgs galaxy spectra were made using the iraf task _ fxcor _ which carries out a fourier cross correlation between the object spectrum and a zero - redshift , zero - velocity dispersion template spectrum .
the task outputs the radial velocity of the object spectrum and the fwhm of the fourier correlation peak between the object and template spectra , which can be converted into a velocity dispersion .
the template spectra used were three velocity standard stars observed with the 6df spectrograph in 2003 march by craig harrison ( hr2574 , hr3145 and hr5888 , which are late g and early k giants ) .
the raw values of the velocity dispersion measured by _ fxcor _ need to be calibrated . to do this each of the standard star spectra
were artificially broadened to a specific velocity dispersion using the iraf task _
gauss_. these artificially broadened spectra were then used to calibrate their ( known ) velocity dispersions against the value measured by _
fxcor_. towards lower velocity dispersions , as one approaches the instrumental velocity resolution ( @xmath33100 km s@xmath12 ) , this calibration becomes less accurate .
however , galaxies with such low velocity dispersions require only very small corrections to their measured lick index values .
consequently , reliable index determinations can still be obtained in low velocity dispersion galaxies .
more importantly , this method of measuring velocity dispersion fails on galaxies that are not well fit by the standard star templates .
this mainly occurs in low signal - to - noise ratio galaxy spectra with young ages and/or strong emission lines .
a comparison of velocity dispersion values derived from the 6dfgs spectra to values in the literature ( hyperleda ) is shown in fig .
[ vdisp_comp ] .
the 6dfgs values show good agreement with the literature values with an average offset of 6.1 kms@xmath12 , and an rms scatter of 20 kms@xmath12 .
both deviation and scatter are greatest at high velocity dispersion , where we find the hyperlead values to be consistently greater than the 6dfgs values .
this is in agreement with bernardi ( 2007 ) who found hyperleda values greater than values derived from the sdss at high velocity dispersion . given the large 6df aperture ( 6.7 arcsec compared to a more typical 2 or 3 arcsec in hyperleda ) it is possible that these deviations are the results of aperture effects .
jrgensen et al .
( 1995 ) investigated the effects of varying the aperture size in observations of velocity dispersions in the inner regions ( i.e. r @xmath33 r@xmath34 ) of early - type galaxies .
their results showed a weak power law dependence of measured velocity dispersions with the size of aperture .
we therefore experimented with correcting velocity dispersions according to jrgensen et al .
( 1995 ) in the sub - sample of 6dfgs galaxies for which hyperlead values were also available . as this subsample of the data is heavily biased to nearby , early - type galaxies , it is well matched to the galaxies used to calibrate the jrgensen et al .
( 1995 ) relation .
the results showed that while the offset was reduced to + 0.75 kms@xmath12 , the scatter was increased to 23 kms@xmath12 .
we also note that the trend for hyperleda values to be greater than values derived from our data remains unaffected by these corrections . as noted above , the jrgensen et al .
( 1995 ) correction is calibrated for the inner regions of early - type galaxies . however , for our 6dfgs sample as a whole , the _ average _ aperture is @xmath01 r@xmath15 ( see fig .
[ sample ] ) .
the sample also contains both early- and late - type galaxies .
this suggests that the corrections presented by jrgensen et al .
( 1995 ) are inappropriate for the 6df sample as a whole .
for such a sample , it is instead informative to consider the study of gebhardt et al .
this analysis of the variation in observed velocity dispersions ( including the effects of rotation ) measured through circular apertures of varying radius in a sample of 26 galaxies of mixed hubble types shows that , within 5 r@xmath15 , the sample exhibits little or no systematic variation of the measured velocity dispersion with aperture size . as a result of the considerations above ,
we have elected to make no corrections for aperture size to the measured velocity dispersion values within this work .
+ to make lick index measurements suitable for direct comparison with single stellar population ( ssp ) models , it is necessary to broaden the spectra to the instrumental resolution of the lick / ids system . the lick / ids fwhm instrumental resolution as a function of wavelength is described in worthey & ottaviani ( 1997 ) where it is shown to vary from @xmath08.5 at the central wavelength ( @xmath05000 ) to @xmath010 at the spectrum ends ( @xmath04000 and 6000 ) . the calculation of the necessary broadening requires the measurement of the 6df instrumental resolution .
this was measured from 6df arc line spectra obtained from the observational time - span of the 6dfgs first data release .
the instrumental resolution of the 6df spectra varies not only with wavelength but also with optical fibre number , so measurements were made of arc lines at various wavelengths , _ for all fibres individually_. a bivariate polynomial was fit to the measured fwhm spectral resolution as a function of both wavelength and fibre number .
the amount of broadening required to match the lick / ids resolution could then be estimated and each 6df spectrum was broadened to this resolution .
the index values given in stellar population models do not include the effect of the velocity dispersion broadening present in galaxy spectra .
it is therefore necessary to introduce a correction to the observed galaxy index values to account for this effect . to measure the velocity dispersion correction required for each lick index , we measured the indices of a range of artificially broadened stellar spectra .
these values of the lick indices measured in the broadened spectra were then compared to those measured in the un - broadened spectra to determine the index corrections for a range of velocity dispersions .
for each index measured in angstroms ( e.g. the fe and ca indices ) a multiplicative correction was defined by : @xmath35 while , for each index measured in magnitudes ( e.g. cn@xmath36 , cn@xmath37 , mg@xmath36 , mg@xmath37 ) , an additive corrections was defined by : @xmath38 where index@xmath39 is the value of the index in the raw stellar spectrum ( i.e. with no broadening ) and index@xmath40 is the index value after broadening of the stellar spectrum to a specific velocity dispersion @xmath41 . for each index
, quadratic lines were fitted to the velocity dispersion versus index correction ratio data .
the index correction required for a particular galaxy could then be determined from the measured velocity dispersion and the quadratic function . estimating errors for spectroscopic measurements is notoriously difficult , mainly because there are several sources of error that can not be well quantified ( e.g. sky - subtraction , poor flux calibration etc ) .
therefore , instead of trying to propagate variance arrays , comparisons of the values from repeated 6dfgs observations of the same galaxy were used to quantify this error .
this method was used for calculating the error in lick indices and velocity dispersions , as well as in h@xmath31 and [ nii ] emission equivalent - widths . clipping .
the solid line is the least squares fit to the bin values .
the dashed line is the relationship between error and signal - to - noise ratio after correcting for the clipping used in forming the rms bins ( see section [ error_measurement ] ) the vertical line marks the lowest signal - to - noise ratio ( 12 @xmath12)used in the present study . *
bottom * : same as top plots except the _ fractional _ error is shown.,title="fig:",width=158 ] clipping .
the solid line is the least squares fit to the bin values .
the dashed line is the relationship between error and signal - to - noise ratio after correcting for the clipping used in forming the rms bins ( see section [ error_measurement ] ) the vertical line marks the lowest signal - to - noise ratio ( 12 @xmath12)used in the present study . *
bottom * : same as top plots except the _ fractional _ error is shown.,title="fig:",width=158 ] clipping .
the solid line is the least squares fit to the bin values .
the dashed line is the relationship between error and signal - to - noise ratio after correcting for the clipping used in forming the rms bins ( see section [ error_measurement ] ) the vertical line marks the lowest signal - to - noise ratio ( 12 @xmath12)used in the present study . *
bottom * : same as top plots except the _ fractional _ error is shown.,title="fig:",width=158 ] clipping .
the solid line is the least squares fit to the bin values .
the dashed line is the relationship between error and signal - to - noise ratio after correcting for the clipping used in forming the rms bins ( see section [ error_measurement ] ) the vertical line marks the lowest signal - to - noise ratio ( 12 @xmath12)used in the present study . *
bottom * : same as top plots except the _ fractional _ error is shown.,title="fig:",width=158 ] in the first data release of the 6dfgs , 3859 galaxies were observed more than once . for all repeat observations
the differences in values were grouped into bins based on the median signal - to - noise ratio of the galaxy spectra .
the signal - to - noise ratio of the lower of the two observations was used for this binning , as this observation would be the cause of most of the error .
if a galaxy had been observed more than twice then the two highest signal - to - noise ratio spectra were used .
the rms of each signal - to - noise ratio bin was determined and used as the @xmath42 value . to remove the biasing effect of large outliers in the bins only the lowest 95.45% of the values ( the @xmath43 limit ) in each bin was used to calculated the rms value .
this requires that the rms be scaled up by 1.164 to take into account this clipping .
a function was fitted to the rms of the bins against the signal - to - noise ratio .
the fitting function used was a modified rectangular hyperbola of the form @xmath44 ( ` a ' and ` b ' the fitted coefficients ) . the error in any measured quantity for a galaxy could then be estimated using the signal - to - noise ratio of the spectrum and the corresponding error from the fitted function .
examples of this procedure can be seen in figure [ error_snr ] .
we note that the majority of the sample with repeat measurements were re - sampled due to the low signal - to - noise of the original observation .
we used the @xmath45-fitting procedure of proctor & sansom ( 2002 ) ( see also proctor et al .
2004a , b and proctor et al . 2005 ) to measure the derived parameters : log(age ) , [ fe / h ] , [ z / h ] and [ e / fe ] ( a proxy for the ` @xmath28'abundance ratio ; see thomas et al .
2003 for details ) .
briefly , the technique for deriving these parameters involves the simultaneous comparison of as many observed indices as possible to models of single stellar populations ( ssps ) .
the best fit is found by minimising the square of the deviations between observations and models in terms of the observational errors ( i.e. @xmath45 ) .
the rationale behind this approach is that , while all indices show some degeneracy with respect to each of the derived parameters , each index does contain _ some _ information regarding each parameter .
in addition , such an approach should be relatively robust with respect to many problems that are commonly experienced in the measurement of spectral indices and their errors .
these include poor flux calibration , poor sky subtraction , poorly constrained velocity dispersions , poor calibration to the lick system and emission - line contamination .
this robustness is of particular importance in the analysis of large numbers of pipe - line reduced spectra such as those of the 6dfgs which can not be flux calibrated and so are not fully calibrated to the lick system .
the method is also relatively robust with respect to the uncertainties in the ssp models used in the interpretation of the measured indices ; e.g. the second parameter effect in horizontal branch morphologies and the uncertainties associated with the asymptotic giant branch .
it was shown in proctor et al .
( 2004a ) and proctor et al .
( 2005 ) that the results derived using the @xmath45 technique are , indeed , significantly more reliable than those based on only a few indices . in order to carry out the comparison of observations to ssp models it was first necessary to select and interpolate the models provided in the literature .
interpolations are required because models in the literature are presented at only five or six discrete metallicities over the @xmath02.5 dex range .
we experimented with a number of different models ( bruzual & charlot 2003 ( hereafter bc03 ) ; thomas , maraston & bender 2003 ; thomas , maraston & korn 2004 and korn , maraston & thomas 2005 ( hereafter kmt05 ) .
the results presented here are those from the models of kmt05 , as these provide the necessary coverage in age ( 0.1 to 15 gyr ) and metallicity ( [ z / h ] from 2.3 to + 0.67 ) , as well as including a detailed modelling of the effects on indices of varying ` @xmath28'abundance ratios .
however , fits were also obtained to the other model sets for comparison . the process by which the best fits were obtained was iterative .
first , fits were obtained for all galaxies using _ all _ the available indices .
the pattern of deviations from the fit so obtained was then used to identify indices that matched the models poorly ( see fig .
[ chis ] ) . due to the problem of emission - line filling of the h@xmath46 , h@xmath47 and h@xmath29 balmer lines in emission galaxies
, we considered ` passive ' and ` emission ' galaxies ( section [ meas_em_lines ] ) separately . in fig .
[ chis ] emission galaxies show the expected large deviation between observed and best fit model values of balmer indices and fe5015 ( which is also emission - line affected ) .
these indices were therefore omitted from the fitting procedure in emission galaxies .
perhaps surprisingly , h@xmath29 and fe5015 also deviate significantly in the ` passive ' galaxies .
we take this to indicate that , despite the rather severe definition of emission galaxies outlined in section [ meas_em_lines ] , at least some of the passive galaxies contain low - level emission that has escaped detection during our emission line measurements . indeed , this was the main motivation for the stringent definition of emission - galaxies .
to minimise the effects of this low - level emission on our age and metallicity estimates , age sensitive h@xmath29 and metallicity sensitive fe5015 indices were also excluded from the fits to passive galaxy data . in both passive and emission galaxies ca4455 was found to fall below best - fit model values .
this effect has been noted in previous studies ( e.g. proctor & sansom 2002 ; clemens et al 2006 ) . the flux - calibration sensitive mg@xmath36 and mg@xmath37 indices were also found to deviate from best fit model values .
this was an anticipated effect , as 6dfgs spectra are only approximately flux calibrated and mg@xmath36 and mg@xmath37 have widely space pseudo - continuum bands .
ca4455 , mg@xmath36 and mg@xmath37 were therefore excluded from the fitting procedure .
fortunately , most of the information lost by the exclusion of the mg@xmath36 and mg@xmath37 indices is captured in the mgb index that is not sensitive to flux - calibration issues .
once poorly fitting indices had been identified and omitted , the fitting of the remaining indices was carried out using a clipping procedure in which indices deviating from the model fit by more than 3@xmath32 were excluded , and the fitting procedure performed again .
this results in an average of less than one additional index per galaxy being clipped from the fits .
most of the clipped indices could be associated with contamination from features such as the 5577 sky - line .
the resultant fits are based on between 10 and 15 indices and show good agreement between observed and best - fit values ( fig .
[ chis ] ) . indeed , the reduced-@xmath45 values given in fig .
[ chis ] ( @xmath48 ) were 1.08 and 1.30 for passive- and emission - galaxies respectively .
this clearly indicates that our data are well represented by the models and that our errors reflect well the differences between observations and models . for each galaxy in the sample , errors in log(age ) and [ z / h ]
were estimated using 50 monte - carlo realisations of the best - fit model indices , assuming the observational errors for the galaxy in question .
we again note that the reduced-@xmath45 values of both passive- and emission - galaxies were close to one , indicating that the assumed errors reflect the average differences between observations and models quite well .
the rms scatter in the log(age ) and [ z / h ] values of the 50 realisations was therefore taken as the individual galaxy error . however , in an effort to better understand the detailed behaviour of the errors , deviations from the best - fit log(age ) and [ z / h ] values of all 50 realisations of each galaxy were retained and are presented in section [ azerr ] . for the purposes of the following analysis
it is important that the derived ages are well defined .
@xmath0 1000 low signal - to - noise galaxies whose error in log(age ) exceeded 0.3 dex were deemed unreliable and were excluded from the analysis .
the average errors for the remaining galaxies were @xmath00.15 dex in both log(age ) and [ z / h ] .
emission lines can be used to separate star - burst and active galactic nuclei ( agn ) galaxies using the [ oiii]@xmath305007/h@xmath29[nii]/h@xmath28 diagnostic .
values of the [ oiii]@xmath305007 , [ nii ] and h@xmath28 emission strengths were estimated from the 6dfgs spectra in a traditional manner ( section [ meas_em_lines ] ) .
the strength of h@xmath29 emission , however , was taken to be the deviation between the measured value of the h@xmath29 index and the best - fit ssp model value .
due to the relatively small observed dynamical range of h@xmath29 in absorption compared to that observed in emission , this is a fairly robust estimate of h@xmath29 emission .
this is evidenced by the classic ` y'-shape evident in the diagnostic plot of [ oiii]@xmath305007/h@xmath29 against [ nii]/h@xmath28 shown in in fig .
[ emission ] .
the position of the line dividing star - burst galaxies from agn varies between studies in the literature . in fig .
[ emission ] we show the lines given by kewley et al .
( 2001 ) ( solid line ) and kauffmann ( 2003a ) ( dashed line ) . in order to ensure reliable identifications of agn we use the definition of kewley et al . in the following .
about 2200 of the 2700 emission galaxies in our sample have reliable estimates of all four of these emission lines , 300 of these lying in the region of the diagram associated with agn as defined by kewley et al .
+ in this section we outline the results of our age and metallicity determinations using lick indices .
the distribution of ages and metallicities are presented in fig .
the sample has been sub - divided into passive- and emission - galaxies as described in section [ meas_em_lines ] .
emission - galaxies classified as agn ( section [ emlingal ] ) are shown as grey symbols , the remainder are plotted in black . before interpreting this plot
, we must note the apparent clusterings and ` zones - of - avoidance ' , the clearest examples of which are evident in the emission galaxy plot .
these are the inevitable result of linear interpolations in non - rectilinear spaces .
returning attention to the actual results of our age and metallicity determinations ; similar trends of increasing [ z / h ] with decreasing log(age ) are evident in both of passive and emission galaxies albeit with emission galaxies tending to younger ages .
galaxies classified as possessing agn in section [ emlingal ] also often possess young central ages .
we investigate the effect of aperture size on derived parameter using fig .
[ aperture ] . in these plots ,
the galaxies with the largest and smallest sizes ( [email protected] arcsec and [email protected] arcsec ) are shown as red and blue points respectively . the blue points therefore represent the galaxies in which the 6.7 arcsec aaomega aperture encompasses the largest fraction of galaxy light ( on average @xmath03.0 r@xmath15 ) , while the red points represent galaxies in which the smallest fraction of galaxy light ( @xmath00.5 r@xmath15 ) is encompassed by the fibre . the observed age
metallicity and age velocity dispersion relations can be seen to be similar in these two extremes , with the most noticeable difference a slightly higher [ z / h ] ( by @xmath00.1 dex ) in large galaxies .
we therefore find aperture - size effects to be generally small . due to the large
, circular aperture of the 6dfgs , a direct comparison with the literature , which are mainly based on long - slit observations of galaxy centres , is not straight forward .
nevertheless , we note that the observed trend is at least qualitatively in agreement with the trends observed in previous studies ( trager et al .
2000 ; proctor & sansom 2002 ; mehlert et al . 2003 ; gallazzi et al .
2005 ; collobert et al .
the young central ages of agn are also consistent with studies in the literature showing that star - bursts are often found to accompany agn activity ( e.g. see cid fernandes et al .
2004 and references therein ) .
an intriguing feature of fig . [ agez ] is the presence of a few old galaxies with extremely low [ z / h ] values , ( @xmath330.5 dex ) .
these galaxies , present in both passive and emission samples , possess a range of velocity dispersions .
visual inspection of both spectra and photometry generally reveals nothing exceptional about these galaxies .
further analysis of these galaxies are beyond the scope of this paper , but we note that their low numbers result in them having no impact on the results and conclusions of the present paper . as in these literature studies , the slope in the age metallicity relation exhibited by our data is similar to the slope of the age
metallicity degeneracy ( worthey 1994 ; ` the 3/2 rule ' ) .
before comparing these results with the photometry , it is therefore clearly important to establish that the age
metallicity degeneracy has been broken . to this end , in the following section we consider the errors in the derived parameters .
as described in section [ errdps ] , errors in age and metallicity were characterised by considering the results of the 50 monte - carlo realisations of each of the 7000 galaxies in our sample .
fits using both the index combinations shown in fig .
[ chis ] ( with and without balmer lines ) were carried out for the purposes of this analysis ( a total of approximately one million realisations ) .
the differences between the input - model values of log(age ) and [ z / h ] and those of each of the realisations were calculated .
these were combined in a number of bins depending upon the age
metallicity of the input models .
the analysis was also carried out using the _ observed _ index values as the inputs to the monte - carlo realisations ( rather than the best - fit model values ) .
no significant quantitative or qualitative differences were found .
[ error ] shows the results of this analysis as 1-sigma confidence contours for three age metallicity bins .
each contour in these plots is based on tens of thousands of individual realisations of the best - fit galaxy data .
the contour levels were defined as the iso - densities corresponding to e@xmath12 of the peak values
( as expected for the 1-sigma contour of a two - dimensional gaussian distribution ) .
the use of the iso - density peak in estimating the extent of 1-sigma confidence contours results in a small uncertainty in the estimates .
however , we estimate these to be of order 2% and therefore insignificant . results of the error analysis for both passive galaxies ( in which balmer indices were included in fits ) and emission galaxies ( in which balmer indices were _ excluded _ from fits ) are shown separately .
we recall that h@xmath29 was excluded from all the fits .
[ error ] also shows the data divided into two signal - to - noise ratio regimes ( s / n@xmath2316 @xmath12 and
s / n@xmath2516 @xmath12 ) . marginal distributions in log(age ) and [ z / h ] are shown at the edges of each main plot .
we note that the positions of the contours in fig .
[ error ] were chosen for clarity only . to get a full picture one must imagine the whole surface of the log(age)[z / h ] plane populated by such contours .
[ error ] shows that , as might be expected , error estimates increase with decreasing signal - to - noise - ratio . for passive galaxies ,
the analysis reveals log(age ) and [ z / h ] errors of @xmath00.1 dex in galaxies with s / n above @xmath016 .
this falls to @xmath00.2 dex in galaxies with s / n below @xmath016 .
the errors of the emission - galaxies ( in which all balmer lines are excluded from the fits ) are somewhat larger , but are generally still of order 0.2 dex .
it is therefore clear that while the age
metallicity degeneracy is still present ( as evidenced by the sloped elliptical error contours in fig .
[ error ] ) , the magnitude of the errors is sufficiently small for its effects to be negligible in our results - even in galaxies fit without balmer lines .
this is the result , and main advantage , of using large numbers ( @xmath2510 ) indices in the determination of ages and metallicities .
the good age resolution achieved is emphasised in the marginal distributions of fig .
[ error ] , in which galaxies older than 10 gyr ( 1.0 dex ) , and galaxies younger than 1.5 gyr ( 0.15 dex ) ( delimited by dashed lines in the marginal distributions ) are largely uncontaminated by galaxies with ages of 3 gyr ( 0.5 dex ) .
since the following analysis concentrates mainly on the very oldest and very youngest galaxies , these provide confidence that results do not suffer significantly from the effects of the age
metallicity degeneracy . the final sample for which velocity dispersions and reliable age / metallicity estimates were measured consists of 4500 passive galaxies and 1000 emission galaxies .
+ the combination of 6dfgs , 2mass and supercosmos data permits the comparison for spectroscopic age and metallicity determinations to the b , r and k band photometric data for some 6000 galaxies .
specifically , by combining the photometry with the velocity dispersion measures from 6dfgs spectroscopy , we are able to investigate trends with age and metallicity in b , r and k band mass - to - light - ratios . this analysis is carried out using a _
dynamical _ mass ( m@xmath51 ) calculated as : @xmath52 where @xmath32 is the central velocity dispersion , r@xmath53 is the half - light radius and the constant c has a value of 5.0 ( cappellari et al .
unfortunately half - light radii are only available for 25% of our sample , while k band 20th magnitude isophotal radii ( r@xmath54 ) are available for the entire sample .
a calibration was therefore carried out by using equation [ masseq ] to calculate masses for the galaxies in the 25% sub - sample ( which also possess much more accurate , aperture corrected , velocity dispersion estimates ) , and comparing them with mass estimates using r@xmath14 ( instead of r@xmath53 ) in the same equation . a plot of the comparison is shown in fig .
[ mass_comp ] .
the correlation has equation m@[email protected] with only 0.12 dex of scatter .
given error estimates in m@xmath51 and m@xmath56 of 0.06 and 0.10 dex respectively , the correlation is clearly extremely good .
masses quoted throughout the remainder of this work are therefore those based on r@xmath14 corrected as detailed . as well as the photometry of 2mass and supercosmos , we also use the photometric results from the ssp models of bruzual & charlot ( 2003 ) .
some key values from the bruzual & charlot models are shown in table [ photometry ] .
we use these models almost exclusively to estimate differential properties ; e.g. the rate of change of mass - to - light - ratio with log(age ) in populations of given [ z / h ] .
it should be noted that the rates of change of the photometric properties are reasonably constant with metallicity and age , as long as age is expressed in logarithmic form .
this clearly indicates that a differential approach is a robust use of the model values .
we shall flag the one occasion in which the models are used in a non - differential manner . .
examples of b , r and k magnitudes from bc03 for single stellar population models of varying age and metallicity are presented .
note that these values are used almost exclusively in a differential manner ( see section [ photo ] ) . [ cols="^,^,^,^,^,^ " , ] we now turn our attention to the effects of the young central populations detected in our 6dfgs spectra on the global properties of galaxies as indicated by their photometry .
there are 400 ` young ' ( @xmath231.5 gyr ) galaxies ( see section [ eamopp ] ) included in this analysis .
plots of mass - to - light - ratio with mass for both young and old galaxies are shown in fig .
mass - binned averages of young and old data are shown as blue and red solid lines respectively .
also shown in these plots as dashed blue lines are model predictions of the effects of the differing ages and metallicities of the young and old populations in each mass bin .
these are based on the comparison of luminosity values from the bc03 ssp models for the differences in age and metallicity of the young and old galaxies in each mass bin .
a direct comparison of the data for young galaxies to these lines would therefore implicitly assume that the young central populations detected in the spectroscopy pervade the galaxies populations _ as a whole _ - i.e represents the same fraction of the galaxy population at all radii .
young galaxies can be seen to be displaced with respect to the old galaxies in a sense consistent with their young ages .
however , the displacements are small compared to the predicted values .
they also become smaller towards higher masses .
the data therefore show that , on average : ( i ) young populations do _ not _ pervade their host galaxies , but are instead centrally concentrated ; ii ) young populations must constitute relatively small fractions of total galaxy masses ; and ( iii ) the mass fraction in the form of a young stellar population must decrease with increasing galaxy mass . to quantify these conclusions , for each mass bin , the fraction of total galaxy stellar mass and luminosity involved in the recent star - burst ( f@xmath57 and f@xmath58 ) were estimated .
these estimates were made by a differential comparison of the bc03 predictions for mass - to - light - ratios to the observed values . in each waveband ,
the difference between the observed mass - to - light - ratios of the young galaxies from those of the old galaxies in the same mass bin ( @xmath59log[m / l]@xmath60 ) was compared to the difference between model predictions for young and old populations of appropriate metallicity ( @xmath59log[m / l]@xmath61 ) . denoting these mass - to - light - ratio differences as s and r respectively ,
we derive : @xmath62_{obs}}-1}{10^{-\delta\log[m / l]_{model}}-1 } \ \ or \ \ \frac{10^{-s}-1}{10^{-r}-1}.\ ] ] the derivation of equation 6 is given in appendix a. we estimate the _ luminosity _ fraction in the recent burst by : @xmath63 finally , the total mass ( m@xmath64 , in solar masses ) is estimated by : @xmath65 where m@xmath66 is the total mass of the galaxy .
the derivation of these expressions again assumes that the young populations at any given mass are seen against a background old population with a metallicity and mass - to - light - ratio appropriate to that mass ( as given by the mass - binned values for old galaxies in table [ mol ] ) .
the young stellar population is nonetheless assumed to _
dominate the central regions _
( such that issues related to degeneracies with burst - strength do not arise ) .
the estimates of mass- and luminosity - fractions and burst - masses are plotted against galaxy mass in fig .
[ mass_frac ] . despite the ( expected ) variation in luminosity fractions between wavebands , agreement between the values for the mass fractions derived from the three wavebands is extremely good .
this suggests that both our assumption of the same underlying dynamical ` tilt ' in young galaxies as exhibited by old galaxies is correct , and that our age and metallicity corrections are accurate .
the trend shows a decreasing mass fraction with increasing mass falling from @xmath010% at 10@xmath67 m@xmath27 to 2% at 10@xmath68 m@xmath27 .
this is in agreement with the study of treu et al .
( 2005 ) , who find the stellar fraction formed in recent times varies from 20%-40% below 10@xmath69 m@xmath27 , to below 1% above 10@xmath68 m@xmath27 .
also shown in fig .
[ mass_frac ] are the estimates of average total mass involved in the recent burst of star formation .
agreement between wavebands is again good .
however , the absolute values given here must be treated with caution , as they are calculated by the only non - differential application of bc03 models used in this work .
nevertheless , in a relative sense , the data exhibit a narrow range of burst - masses and an apparent _ upper limit _ on their size in high mass galaxies . as a final consideration
we next investigate the implications of our findings for colour - magnitude relations .
note that these relations are independent of dynamical effects . in fig .
[ cmr ] we plot the colour - magnitude relations b k and r k against m@xmath70 .
galaxies with both old ( grey points ) and young ( black points ) central populations are shown .
averages ( binned by luminosity ) of the old and young populations are again shown as solid red and blue lines respectively .
slopes in the old galaxies are in good accord with literature values ( thin black lines in each plot ) . as the sample includes late - type galaxies , particular amongst the young galaxies , these must lie close to the ` red sequence ' in our colour - magnitude diagrams .
this suggests they therefore correspond to the ` dusty red - sequence ' of wolf , gray & meisenheimer ( 2005 ) .
the lack of galaxies in the ` blue - cloud ' is almost certainly the result of a number of selection effects .
these include the effects of template mismatches during velocity dispersion measurements and the removal of galaxies with poor fits or large errors during age / metallicity measurement .
[ cmr ] shows that , for both colours , young galaxies tend to be bluer than the old galaxies .
however , the offsets are extremely small . our results from the previous section indicate that this is largely the result of the low mass fractions formed in the recent star formation events .
[ cmr ] therefore shows that galaxies lieing on , or close to , the red sequence may not be ` red and dead ' , but may still be forming modest numbers of new stars . indeed , we note that _ most _ of the galaxies with young central populations also exhibit on - going star formation ( fig .
[ agez ] ) .
we have measured ages and metallicities from the 6dfgs spectra of @xmath07000 galaxies of all morphological and emission types .
we have demonstrated that the age
metallicity degeneracy has been broken in this study , particularly with regard to the very young and very old galaxies used in the subsequent analysis .
these data for the _ central regions _ are compared with the _ global _ photometry from 2mass and supercosmos for the @xmath06000 galaxies for which such photometry is available . combining spectroscopy and photometry we are able to investigate trends in global b , r and k band mass - to - light - ratios with the central stellar populations of the galaxies .
an age - metallicity trend is identified in passive galaxies that is consistent with the results of similar studies in the literature .
the stellar populations of galaxies exhibiting emission lines ( arising from either star - bursts or agn ) follow a similar trend , but exhibit younger ages than their passive counterparts . to minimise the effects of complex mixtures of stellar populations , we confine our analysis to the data for very old or very young galaxies .
we find a steep mass - metallicity relation in old galaxies ( logarithmic slope of @xmath00.25 ) , while in young galaxies the trend is consistent with zero slope .
using only the old galaxies , we find a dynamical trend in mass - to - light - ratios with mass of logarithmic slope @xmath00.15 .
this is in good agreement with values reported in the literature .
however , there is no agreement in the literature as to the cause of this tilt .
padmanabhan et al .
( 2004 ) concluded that most of the tilt is due to variations in dark matter content . from the modelling perspective ,
robertson et al . (
2006 ) perform simulations of a series of mergers and investigate the scaling relations of their merger remnants . from mergers of gas - rich disk galaxies
they derive a fp with @xmath28 = 1.55 and @xmath29 = 0.82 , in good agreement with our k band values ( see table 3 ) . in their models
the tilt away from the virial scaling relation is caused by gas dissipative effects and not dark matter content .
trujillo et al . ( 2004 ) on the other hand attribute the tilt to ` non - homology ' , deriving almost identical contributions to the tilt from dynamical and stellar population effects as those we find in this work .
our findings , however , also show that whatever the cause of the tilt , it is not a simple in mass - to - light - ratio ( or its scatter ) with mass or luminosity .
instead we find a that a more complex description will be required ; this will be a subject of a future paper .
considering extremely young galaxies , our results show that the properties of the central stellar populations identified by the spectroscopy do _ not _ generally correlate with the global properties of their host galaxies .
indeed , the data indicate that recent bursts of star formation are limited to a mass of @xmath3310@xmath24 m@xmath27 and are strongly centrally concentrated .
consequently , in high mass galaxies exhibiting young central populations , recent star forming events have negligible effect on their global photometry . as a consequence
we have shown that galaxies sitting on , or near , the red sequence may not be ` red and dead ' , but may still be forming modest numbers of new stars .
we therefore conclude that the young ages found in studies of galaxy centres in this and other spectroscopic studies should not be considered representative of the age of their host galaxies as a whole .
they may , however , be reasonably interpreted as indicators of recent assembly events . as a result of these considerations ,
the small scatter in photometric scaling relations , which initially motivated the monolithic collapse models , ceases to be inconsistent with either the hierarchical merging model or the results of the lick index studies indicating young central populations . in future work
we will include the results from the completed 6df galaxy survey and report on the full chemical properties ( including @xmath28-abundance ratios ) of the central regions of galaxies .
the effective radii , peculiar velocities and environmental parameters of the 6dfgs sample , currently being analysed by the 6dfgs team , will be included .
this will allow not only more accurate determination of photometric properties and a more detailed analysis of the fundamental plane , but also the investigation of the strong _ quantitative _ predictions of trends in galaxy formation / assembly epochs and metals - content with galaxy mass and environment . +
* acknowledgements * this publication makes use of data products from the two micron all sky survey ( 2mass ) , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
the authors acknowledge the data analysis facilities provided by iraf , which is distributed by the national optical astronomy observatories and operated by aura , inc .
, under cooperative agreement with the national science foundation .
we thank the australian research council for funding that supported this work .
the measuring of line values from 6dfgs spectra was begun in philip lah s honours thesis 2003 `` exploring the stellar population of early - type galaxies in the 6df galaxy survey '' which looked at the 6dfgs early data release , ( december 2002 ) .
this honours thesis was supervised by matthew colless and heath jones .
assistance with the software used in this work was provided by craig harrison .
lachlan campbell provided additional assistance with the 6dfgs data particularly with the velocity dispersions and 6df arc spectra .
the authors also thank chris blake for valuable input . * references * + baldry i.k .
et al . , 2002 ,
apj , 569 , 582 + baum w.a . , 1959 , iau symp .
10 , the hertzsprung - russell diagram , ed .
j. l. greenstein ( paris : iau ) , 23 + bernardi m. et al .
, 2003a , aj , 125 , 1882 + bernardi m. et al .
, 2003b , aj , 125 , 1866 + bernardi m. et al . , 2006 ,
aj , 131 , 2018 + bernardi m. , 2007 , aj , 133 , 1954 + blumenthal g.r . ,
faber s.m .
, primack j.r .
, rees m.j . , 1984 ,
nature , 311 , 517 + bower r.g .
, lucy j.r . , ellis r.s . , 1992a , mnras , 254 , 589 + bower r.g . , lucy j.r .
, ellis r.s .
, 1992b , mnras , 254 , 601 + bruzual a.g .
, charlot s. , 2003 , mnras , 344 , 1000 + caldwell n. , rose j.a .
, concannon k.d . , 2003 , aj , 125 , 2891 + cappellari m. et al . , 2006 , mnras , 366 , 1126 + carlberg r.g . , 1984 ,
apj , 286 , 403 + chang r. , gallazzi a. , kauffmann g. , charlot s. , ivezi z. , brinchmann j. , heckman t.m . , 2006 , mnras , 366 , 717 + cid fernandes r. , gu q. , melnick j. , terlevich e. , terlevich r. , kunth d. , rodrigues lacerda r. , joguet b. , 2004 , mnras , 355 , 273 + cid fernandes r , mateus , a. , sodr l. , stasinska g. , gomes j.m . , 2005 , mnras , 358 , 363 + clemens m.s . , bressan a. , nikolic b. , alexander p. , annibali f. , rampazzo r. , 2006 , mnras , 370 , 702 + colless m. , saglia r.p . , burstein d. , davies r.l . , mcmahan jr r.k . , wegner g. , 2001a , mnras , 321 , 277 + colless m. et al . ,
2001b , mnras , 328 , 1039 + collobert m. , sarzi m. , davies r.l .
, kuntschner h. , colless m. , 2006 , mnras , 370 , 1213 + de lucia g. , springel v. , white s.d.m . , croton d. , kauffmann g. , 2006 , mnras , 366 , 499 + dressler a. , lynden - ball d. , burstein d. , faber s.m . , terlevich r. , wegner g. , 1987 , apj , 313 , 42 + eisenstein d.j .
et al . , 2003 ,
apj , 585 , 694 + fernandes r.c .
, mateus , a. , sodr l. , stasinska g. , gomes j.m .
, 2005 , mnras , 358 , 363 + gallazzi a. , charlot s. , brinchmann j. , white s.d.m .
, tremonti c. , 2005 , mnras , 362 , 41 + gebhardt , k. et al . ,
apj , 539 , l13 + gibbons r.a .
, fruchter a.s . ,
bothun g.d .
, 2001 , aj , 121 , 649 + gonzlez j. j. , 1993 , phd thesis , univ .
california + hambly n.c .
, irwin m.j . , macgillivray h.t . , 2001 ,
mnras , 326 , 1295 + hudson m.j . , lucey j.r . ,
smith r.j . , steel j. , 1997 , mnras , 291 , 488 + jarrett t.h . , chester t. , cutri r. , schneider s. , skrutskie m. , huchra j.p .
, 2000a , aj , 119 , 2498 + jarrett t.h . ,
cheste , t. , cutri r. , schneider s. , rosenberg j. , huchra j.p . , mader j. , 2000b , aj , 120 , 298 + jones d.h .
et al . , 2004 ,
mnras , 355,747 + jones d.h . , peterson b.a . , colless m. , saunders w. , 2006 , mnras , 369 , 25 + jrgensen i. , franx m. , kjrgaard p. , 1995 ,
mnras , 276 , 1341 + jrgensen i. , franx m. , kjrgaard p. , 1996 ,
mnras , 280 , 167 + kauffmann g. , 1996 , mnras , 281 , 487 + kauffmann g. et al . , 2003 , mnras , 346 , 1055 + kauffmann g. et al . ,
2003a , mnras , 341 , 33 + kauffmann g. et al .
, 2003b , mnras , 341 , 54 + kewley l.j .
, dopita m.a .
, sutherland r.s . ,
heisler c.a . , trevena j. , 2001 , apj , 556 , 121 + korn a.j . , maraston c. , thomas d. , 2005 , a&a , 438 , 685 + larson r.b . , 1974 , mnras , 166 , 585 + lewis i. et al . , 2002 , mnras , 334 , 673 + mehlert d. , thomas d. , saglia r.p
, bender r. , wegner g. , 2003 , a&a , 407 , 423 + moore s.a.w .
, 2001 , ph.d thesis , univ of durham + padmanabhan n. et al . , 2004 , new astron .
, 9 , 329 + pahre m.a . , djorgovski s.g .
, de carvalho r.r . , 1998 , aj , 116 , 1591 + proctor r.n . , sansom a.e . , 2002 , mnras , 333 , 517 + proctor r.n . , forbes d.a
. , beasley m.a .
, 2004a , mnras , 355 , 1327 + proctor r.n .
, forbes d.a . ,
hau g.k.t . ,
beasley m.a . ,
de silva g.m . , contreras r. , terlevich a.i . , 2004b , mnras , 349 , 1381 + proctor r.n . , forbes d.a
, forestell a. , gebhardt k. , 2005 , mnras , 362 , 857 + robertson b. , cox t.j . , hernquist l. , franx m. , hopkins p.f . , martini p. , springel v. , 2006 , apj , 641 , 21 + smith r.j .
, hudson m.j . , lucey , j.r . ,
nelan , j.e . ,
wegner , g.a .
, 2006 , mnras , 369 , 1419 + spergel d.n .
et al . , 2003 ,
apjs , 148 , 175 + terlevich a.i .
, forbes d.a . , 2002 , mnras , 330 , 547 + thomas d. , maraston c. , bender r. , 2003 , mnras , 339 , 897 + thomas d. , maraston c. , bender r. , mendes de oliveira c. , 2005 , apj , 621 , 673 + thomas d. , maraston c. , korn a. , 2004 , mnras , 351 , 19 + trager , s. , faber , s. , worthey , g. , gonzalez , j. , 2000 , aj , 120 , 165 + tremonti c. et al .
, apj , 613 , 898 + treu t. , ellis r.s . , liao t.x . , van dokkum p.g . ,
tozzi p. , coli , a. , newman j. , cooper m.c .
, davis m. , 2005 , apj , 633 , 174 + trujjillo i. , burkert a. , bell e.f . , 2004 ,
aj , 600 , l39 + white s.d.m . ,
rees m.j . , 1978 , mnras , 183 , 341 + wolf c. , gray m.e . , meisenheimer k. , 2005 , a&a , 443 , 435 + worthey g. , 1994 , apjs , 95,107 + worthey g. , ottaviani d.l . , 1997 , apjs , 111 , 377 + york d.g .
et al . , 2000 ,
aj , 120 , 1579 + [ lastpage ] | we present an analysis of the stellar populations as a function of mass in a sample of @xmath07000 galaxies of all morphological and emission types from the 6df galaxy survey ( 6dfgs ) .
we measure velocity dispersions and lick indices from the spectra of the central regions of these galaxies , deriving ages and metallicities from the lick indices using stellar population models .
we also derive dynamical masses and dynamical mass - to - light ratios for these galaxies by combining the central velocity dispersions with global photometry in the b , r and k bands from supercosmos and 2mass .
together , these data allow us to reduce the degeneracies between age , metallicity and star formation burst - strength that have limited previous studies .
we find that old galaxies exhibit a mass - metallicity relation with logarithmic slope d[fe
/ h]/dlog@xmath1 , while young galaxies show slopes consistent with zero .
when we account for the effects of the mass - metallicity relation , we obtain a single , consistent relation between mass - to - light ratio and mass for old galaxies in all passbands , @xmath2 . as we have accounted for stellar population effects , this remaining variation in the mass - to - light with mass ( the residual ` tilt ' of the fundamental plane ) must have a dynamical origin .
however , we demonstrate that any simple trend between mass - to - light - ratio and mass or luminosity is inconsistent with the observations , and that a more complex relationship must exist .
we find that the central regions of galaxies of all masses often exhibit young stellar populations .
however it is only in the lowest - mass galaxies studied ( @xmath3 ) that these populations are evident in the global photometry . in higher - mass galaxies , young central populations
have decreasing influence on the global photometry , with there being no discernible impact in galaxies more massive than @xmath4 .
we conclude that the young stellar populations detected in spectroscopic studies are generally centrally concentrated , and that there is an upper limit on the mass of star - forming events in massive galaxies .
these results have important ramifications for mass - to - light ratios estimated from photometric observations .
latexl-.36em.3ex-.15em t-.1667em.7ex-.125emx [ section ] [ firstpage ] galaxies : general , galaxies : stellar content , galaxies : kinematics and dynamics |
the use of two laser fields to dress the eigenstates of a three - level system has attracted significant attention over past two decades @xcite .
the interaction of light with a coherent atomic ensemble leads to important quantum interference phenomena , including electromagnetically induced transparency ( eit ) @xcite , in which the laser - induced coherence of atomic states leads to quantum interference between the excitation pathways that control the optical response .
it can be used to manipulate the linear and nonlinear susceptibilities of matter from the near - infrared to ultraviolet .
this has resulted in applications such as slowed and stored light as well as nonlinear frequency conversion @xcite .
presently there is great interest in the application of light in the x - ray regime , produced by high - order harmonics and free electron lasers , to investigate novel coherent x - ray optical phenomena . in this context
, ultrafast electron dynamics of the autoionization processes in atoms has gained attention in recent years .
the advent of femtosecond and attosecond xuv laser pulses in a pump - probe setting has made it possible to follow these processes in real time @xcite .
these experiments involving two coherent laser pulses allow study of the time - resolved interaction of the direct ionization pathway with pathways involving one or more resonant states .
this opens the possibility of new ways to monitor and control the autoionization dynamics @xcite and consequently the xuv absorption properties of the dressed medium .
theoretical studies due to @xcite and @xcite predicted the transformation of the original unperturbed fano line - shape @xcite into an autler - townes doublet due to rabi flopping between a pair of coherently - coupled doubly - excited states .
@xcite observed eit - like behavior in an xuv probe induced by coherent coupling of the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states in helium , where the xuv probe was created from ir laser - produced high - order harmonics .
specifically , that study used a femtosecond infrared ( ir ) pulse both to couple the doubly - excited states and to produce the probe pulse .
@xcite predicted an eit for x - rays in laser dressed neon .
more recently , @xcite used a two - color multiphoton ionization of helium in combined ir laser and xuv high harmonic fields to control the transparency .
they showed that this scheme can induce full electromagnetic transparency .
quantum mechanical calculations of the photoabsorption spectra for the dressed atoms represent a challenge for theory , especially when the dressing laser pulse couples doubly - excited states .
the time - consuming solution of the time - dependent schrdinger equation ( tdse ) is required for every photon energy of the probe pulse .
an alternative theoretical treatment has been introduced by @xcite and successfully used in ref .
it is based on the floquet representation and on the calculation of the autocorrelation function that is indepdendent of the probe laser photon energy .
therefore , the time and memory demanding numerical calculation of the autocorrelation function is performed only once . since the photoabsorption cross section has a simple analytical dependence on the autocorrelation function , its calculation is not difficult .
however , the approach introduced in ref .
@xcite involves a diagonalization of the hamiltonian matrix in the floquet representation , typically having dimensions @xmath6 and this step is still challenging for a supercomputer . to our knowledge , there is no method which enables a routine theoretical calculation of the photoabsorption spectra in the presence of a strong dressing laser pulse .
the aim of the present work is to make a step towards the improvement of the theoretical treatment of this problem .
this study deals with the modified fano line - shape in the photoabsorption spectrum of the he 2s2p ( @xmath0)doubly - excited state caused by the coupling of this state with the 2p@xmath1 @xmath2doubly - excited state by an ir femtosecond pulse .
@xcite studied this effect experimentally and interpreted obtained results as eit - like behavior .
a simplified three - level@xcite model treatment of this problem was included in ref .
the model describes formation of an autler - townes doublet and takes into account the effect of the optical field ionization of the doubly - excited states due to the ir dressing laser using a somehow _ ad hoc _ adaptation of the ammosov - delone - krainov ( adk ) method .
our approach presented in this work is based on the numerical solution of the tdse .
it takes into account the possible coupling of more than two autoionizing states and the continuum in principle more accurately than the approach presented in ref .
therefore , the present treatment of the optical field ionization of the doubly - exited states due to the coupling laser field is more quantitative as well .
photoabsorption spectra calculated using parameters corresponding to the experimental setup used in ref .
@xcite are compared with results obtained for reduced ir pulse intensity and for significantly reduced ir detuning .
comparison of all these results allows us to study the effect of the optical field ionization and of the ir laser detuning on the formation of eit - like structures in the spectra .
the rest of this paper is organized as follows : the theoretical approach and assumptions of our model are discussed in sec .
[ sec : theo ] .
a detailed description of the field free hamiltonian and static photoabsorption spectrum is given in sec .
[ sec : staticfree ] , while sec .
[ sec : dressed ] discusses the photoabsorption yields of the dressed atom .
atomic units are used throughout the rest of the paper .
the model introduced in the present study is based on the experimental publication of to @xcite . in that experiment
the 2s2p ( @xmath0)(@xmath7p@xmath8 ) and 2p@xmath1 @xmath2(@xmath7s@xmath9 ) doubly - excited states of he are coupled by a strong 800 nm field .
the 2s2p ( @xmath0)and 2p@xmath1 @xmath2states are located at 60.15 @xcite and 62.06ev @xcite above the 1s@xmath1ground state , respectively .
this system is then probed after a variable time delay by an xuv pulse .
the xuv probe pulse is produced with a table - top , laser - based setup by high - order harmonic generation ( hhg ) .
the authors use a commercial ti : sapphire laser system ( 2.4w , 800 nm , 42fs , 1khz ) to produce the optical dressing and hhg beam .
the estimated photon flux at the source is @xmath10 photons per pulse for the high - order harmonic centered at 60.2ev .
the xuv pulse duration is estimated to be 30fs fwhm .
the dressing pulse peak intensity is @xmath11w / cm@xmath1 and its duration is 42fs fwhm .
the spectrometer used in @xcite measures the transmission spectrum instead of the absorption spectrum .
the unknown fraction of dressed atoms in the interaction volume prevents the photoabsorption spectrum of the dressed atom from being retrieved experimentally .
therefore , authors measure the transient absorption spectra and identify eit - like structures . the xuv laser pulse couples the 1s@xmath1ground state with the 2s2p ( @xmath0)autoionizing state with energy 35.56ev above the @xmath13 ionization threshold .
therefore , it is reasonable to assume that every absorption of an xuv photon leads to photoionization .
the aim of the present work is to calculate the population of the ground state at the end of the pulse as a function of the xuv photon energy @xmath14 for different dressing - probe time delays @xmath15 .
the computed loss of the ground state population is interpreted as the total photoabsorption yield .
its comparison with the yields obtained when the ir pulse is absent allows for recognition of eit - like structures .
helium is a two - electron system .
the two - particle basis set used in the present calculations consists of @xmath16-coupled independent - particle basis functions @xcite .
the hamiltonian for the interaction of each electron with the core is @xmath17 where @xmath18 is the coordinate of the electron with respect to the nucleus .
the corresponding ( radially rescaled ) single - particle radial wave function @xmath19 for an electron with orbital angular momentum @xmath20 , is given by @xmath21 the atomic hamiltonian including the interaction between the electrons is @xmath22 where @xmath23 is the distance between the electrons . to calculate the time development of the ground state population we solve the tdse for the two linearly - polarized laser fields in the length gauge : @xmath24\psi(\mathbf{r}_1,\mathbf{r}_2,t ) , \label{eq : alltdse}\ ] ] where @xmath25 are the interaction operators for the ir and xuv laser pulse in the dipole approximation , respectively .
@xmath26 denotes the complex absorbing potential ( cap ) added to treat ionization properly within the context of a finite volume calculation . as in the experimental convention
, a positive time delay @xmath15 means that the xuv probe pulse center arrives later than the pump pulse center .
the electric fields @xmath27 and @xmath28 are given by @xmath29 where @xmath30 are the ir and xuv laser field frequencies , @xmath31 is the unit vector along the polarization axis , @xmath32 are the peak amplitudes of the electric fields and @xmath33 are the pulse envelopes . throughout this work
it is assumed that both envelopes have the @xmath34 form @xmath35 resp .
, @xmath36 where @xmath37 denote the durations of the respective laser pulses . the second term in eq .
( [ eq : irenvelop ] ) denotes the ir postpulse which appears in the experiment @xcite with time delay @xmath38fs and duration @xmath39fs .
the cap used in the present calculations has the form @xcite @xmath40 , & \text{otherwise , } \end{array } \right .
\label{eq : capop}\ ] ] where @xmath41 is the strength of the cap and @xmath42 is the inner boundary of the absorber .
the tdse ( [ eq : alltdse ] ) is integrated in a box using the expansion of the time - dependent wave function @xmath43 in a basis set consisting of eigenstates of the field - free atomic hamiltonian @xmath44 : @xmath45 where @xmath46 denotes the n^th^ eigenfunction of @xmath44 having total angular momentum @xmath47 and corresponding to the real `` box '' eigenenergy @xmath48 : @xmath49 the expansion ( [ eq : timexpand ] ) includes @xmath50 eigenstates for the angular momentum @xmath47 , and it includes all the angular momenta up to @xmath51 . the basis set used in the variational calculations of the field - free eigenstates @xmath46 consists of _ ls_-coupled independent - particle basis functions .
antisymmetric two - electron basis functions coupled to form a state of definite @xmath52 , @xmath47 , and parity @xmath53 are constructed in terms of one - electron orbitals ( [ eq : oneorbs ] ) as explained in refs .
the radial one - electron wave functions @xmath54 are obtained by solving eq .
( [ eq : oneorbs ] ) in a radial box of size @xmath55 with boundary conditions @xmath56 for both positive and negative energies @xmath57 . the index @xmath58 denotes the number of nodes of the radial wave function inside the box .
( [ eq : oneorbs ] ) is solved by expansion of the radial orbitals in the basis set of _ b_-splines @xcite @xmath59 satisfying the bound - state boundary conditions : @xmath60 the radial box is divided into several intervals ( denoted by index @xmath61 ) by knots and the _ b_-splines are defined in terms of the knots in each interval .
use of the _
ls_-coupled basis as discussed above is nowadays a standard technique frequently employed in the theoretical treatment of both time - dependent @xcite and time - independent @xcite problems for two active electron atomic systems .
the radial one - electron orbitals are often expressed in terms of _
b_-splines @xcite or discrete variable representation ( dvr ) basis sets @xcite . in order to suppress the reflections from the outer boundary of the radial box in the solution of the tdse ( [ eq : alltdse ] ) and to keep its size reasonably small
, the cap term ( [ eq : capop ] ) is added to the hamiltonian used for the time - propagation .
the presence of this term makes the total hamiltonian of the system a complex symmetric rather than hermitian operator @xcite .
one note should be made at this point regarding the form of the matrix elements of all the operators used here . since
all the calculations are performed in the eigenrepresentation of the field - free atomic hamiltonian @xmath44 not including the cap , the operators @xmath44 , @xmath62 and @xmath63 have real matrix elements , as it is convenient in hermitian quantum mechanics .
@xmath26 is the only non - hermitian term in the total hamiltonian .
therefore , it is not necessary to distinguish the standard hermitian scalar product and complex symmetric product throughout the present work ( see ref .
@xcite and references therein for details ) .
both laser pulses are linearly polarized along the @xmath64-axis , and of course the @xmath3 ground state of he has zero projection @xmath65 of the total orbital angular momentum on the polarization axis .
accordingly it is also enough to consider only the @xmath64-component of the dipole operator . according to the selection rules for the matrix elements of the dipole moment operators @xmath66 , these are non - zero only if @xmath67 @xcite .
projection of the tdse ( [ eq : alltdse ] ) on the eigenstates of the field - free hamiltonian ( [ eq : eigstates ] ) yields the following set of coupled differential equations of the first order : @xmath68\mathbf{c}(t ) , \label{eq : tdmat}\ ] ] where @xmath69 is the vector of expansion coefficients used in ( [ eq : timexpand ] ) and the matrices on the right hand side are the representations of the corresponding operators in the basis set @xmath46 introduced above .
the dimension of this system of differential equations is @xmath70 . in the present representation
@xmath71 is a real diagonal matrix of the field - free atomic eigenenergies . since @xmath26 given by eq .
( [ eq : capop ] ) has no angular dependence , @xmath72 is an imaginary and symmetric block diagonal matrix with @xmath73 non - zero blocks having dimensions @xmath50 .
@xmath74 are real symmetric banded matrices with only non - zero blocks corresponding to the coupling of the total angular momentum @xmath47 with @xmath75 , because we only consider @xmath76 states in this study .
the physical picture of the phenomenon studied here suggests that the ir dressing laser pulse couples the singlet autoionizing states 2s2p ( @xmath0 ) , 2p@xmath1 @xmath2and the continuum states made accessible by the ir pulse .
the energy difference between the ground state and the lowest odd - parity , singlet excited state - 1s2p(@xmath7p@xmath77 ) of the field - free atom is so high that it requires more than 18 photons to be absorbed to excite the ground state by the ir pulse .
transitions of higher orders would thus be necessary to excite higher excited states directly from the ir pulse acting on the ground state .
therefore , it is a reasonable assumption to neglect the ir coupling of the ground state with excited states .
note that the weak effect of the ir field for intensities considered in this work is also suggested in ref .
therefore , the matrix elements of the dipole operator in the interaction matrix for the ir field @xmath78 were explicitly set to zero ( considering the ground state the first element in the basis set ) .
this assumption allows for a dramatic improvement in the convergence of the solutions of the tdse with respect to the size of the basis set . on the other hand , it is the primary role of the xuv probe laser pulse to couple the ground state with the autoionizing state 2s2p ( @xmath0)and with nearby continuum states . since it has rather high photon energy and low intensity ( perturbative regime ) , it is unlikely that the xuv coupling between the doubly - excited states and the continuum states will play any significant role .
therefore , in the present study all the dipole matrix elements in the xuv interaction matrix @xmath79 except those coupling to the ground state @xmath80 , are explicitly set to zero .
note that a similar approximation for the xuv coupling in a similar context has been employed in ref .
@xcite . using this approximation the system of
coupled differential equations ( [ eq : tdmat ] ) can be written as follows : @xmath81\mathbf{c}(t ) .
\label{eq : approxmat}\ ] ] the structure of the matrices is showed in fig .
[ fig : eqstr ] . ) ] .
the grey areas denote the non - zero blocks corresponding to definite total angular momenta and their dipole coupling .
matrix @xmath82 has zero first line and first column , while matrix @xmath83 has non - zero only the first row and column .
[ fig : eqstr ] the atomic system is initially in the ground state . considering the ground state to be the first element of the basis set , the initial condition can be written as @xmath84 .
the neglect of the ground state coupling with the other field - free eigenstates via the ir laser pulse allows for reduction of the total interval for the propagation of the tdse as well . since @xmath85 , it would be necessary to start the propagation of the tdse ( [ eq : tdmat ] ) in the time @xmath86 , if this assumption were not made
. it would also then be necessary to perform the propagation of the wave function until the time @xmath87 .
although the total wave function described by the vector @xmath69 changes at times @xmath88 due to the presence of the cap , the population of the ground state will remain constant after the end of the pulses , since the ground state wave function is well localized in the region where the cap is zero .
therefore , for the purpose of the present project it is enough to perform the propagation until @xmath89 and extract the population of the ground state @xmath90 .
the approximation that the ir field does not affect the ground state allows for the start of the propagation at time @xmath91 and terminate it at time @xmath92 . in view of the long duration of both pulses , this is an important simplification of the numerical calculations .
the method of solution of the tdse used in the present work is based on the split operator method in the basis set representation and it is described in the appendix in detail .
this represents an alternative approach to the higher - order quadrature methods @xcite frequently used in context of the interaction of atoms and molecules with strong laser fields .
application of the split - operator technique is straightforward in case of the wave - packet propagation ( for details see ref .
@xcite and references therein ) .
however , the exponential of the time dependent operator remains a challenge in the basis set representation .
@xcite introduced an implementation in the dvr basis set which has properties that allow for an efficient numerical calculation of these exponentials . in the implementation
used here these time - dependent exponentials are calculated in the diagonal representation of the dipole operators , as is discussed in the appendix .
the photoabsorption yields obtained from the solutions of the tdse ( [ eq : approxmat ] ) together with the static photoabsorption yield enable a calculation of the transient absorption spectra .
this allows to relate the theoretical results presented here with the experimental transient absorption spectra published in ref .
although we have found that the eigenrepresentation of the atomic hamiltonian @xmath44 without the cap is more practical for the solutions of the tdse ( [ eq : approxmat ] ) , it is useful to study the spectrum of the complex symmetric hamiltonian @xmath93 .
this representation allows for easier optimization of the cap parameter @xmath94 and of the size of the radial box @xmath95 .
the complex eigenenergies allow for a better insight into the structure of the discretized continuum and evaluate its quality for the solution of the full tdse as well .
the eigenfunctions @xmath96 of the complex symmetric operator @xmath93 are given by the equation @xmath97 where the indices @xmath98 have identical meaning as in eq .
( [ eq : eigstates ] ) .
the complex eigenergies can be written as @xmath99 .
they should reproduce well the energy positions @xmath100 and widths @xmath101 of the 2s2p ( @xmath0)and 2p@xmath1 @xmath2autoionizing states for @xmath102 .
in addition , the density of the discretized continuum states must be sufficient to accurately describe the interaction between the autoionizing states and the continuum and to allow for converged calculation of the photoionization yield .
these properties of the basis set are controlled by the size of the box @xmath95 , the position of the absorbing boundary @xmath42 and the strength of the cap @xmath94 , as is described in ref .
these parameters have been optimized in the present calculations to the values @xmath103a.u . and @xmath104a.u .
these values allow for a good representation of the complex eigenenergies of the autoionizing states , they yield negligible decay widths of the ground and low excited states , and they provide a sufficient number of continuum states above the ionization threshold . the radial box is spanned by 700 _ b_-splines of the sixth order .
ls_-coupled expansion includes the single - particle partial waves up to @xmath105 , every partial wave includes 200 radial orbitals ( see eq .
( [ eq : oneorbs ] ) ) . in order to describe the autoionizing states of the interest properly ,
the ci expansions of the states with @xmath106 and @xmath107 include the doubly - excited configurations , where both electrons can occupy orbitals with @xmath108 and @xmath109 . in the remaining configurations one electron
is restricted to occupy a 1s , 2s , or 2p orbital , while the other electron can occupy any orbital , constrained so that the configurations contribute to the states with given spin @xmath52 , total angular momentum @xmath47 and parity @xmath53 .
the higher partial waves @xmath20 included in the expansion are required for the representation of the continuum states having high total angular momentum @xmath47 ; they are essential for obtaining converged solutions of the full tdse including the nonperturbative ir laser pulse .
optimization of the parameter @xmath94 as is described in the ref .
@xcite using this basis set yields the value @xmath110 .
the complex energies of the autoionizing states calculated using the basis set presented here are compared with previously published experimental results @xcite in table [ tab : respos ] .
@dcdcd & & + & & & & + & 59.687 ( 59.688 ) & 60.15 & 0.0411 ( 0.0456 ) & 0.0373 + & 61.747 & 62.06 & 0.0063 & 0.0059 + the widths of both autoionizing states calculated here are in good agreement with the experimental results .
the discrepancies in their relative positions with respect to the ground state are mainly due to the inaccurate representation of the 1s@xmath1ground state by our truncated atomic hamiltonian @xmath71 .
it gives a computed ground state energy of @xmath111a.u .
, while the accurate experimental value is @xmath112a.u .
@xcite . to improve the ground state energy ( and subsequently the relative positions of the double excited states )
a significantly larger basis set would be necessary .
however , the difference of doubly - excited state energies is more important for the calculations of the photoabsorption spectra than their relative position with respect to the ground state . as can be seen from table [ tab : respos ] , the energy difference according to the calculations is @xmath113ev , while the experimental results give @xmath114ev . the distribution of the @xmath7s@xmath115 and @xmath7p@xmath77 continuum states in the complex energy domain is plotted in fig .
[ fig : constat ] . the ir laser pulse assumed here
allows for the optical field ionization of the doubly - excited states into the @xmath116 continuum .
since @xcite suggested that inclusion of this effect is important for the proper treatment of the phenomenon studied here , it is important to have a sufficiently high density of continuum states with higher total angular momenta @xmath47 in the energy range above the @xmath116 ionization threshold . spanning the interval of energies probed by the xuv laser pulse .
energies of states with the total angular momentum @xmath106 and @xmath107 are plotted .
the autoionizing states are labeled by their principal independent particle ci configurations .
the autoionizing state with energy 62.22ev without label does not have a single dominant ci configuration , its discrete component is a linear combination of the 2s3p and 3s2p configurations .
the @xmath116 ionization threshold is at energy 64.86ev . ]
present calculations of the dipole matrix @xmath82 yields @xmath117a.u . for the matrix element coupling the 2s2p ( @xmath0)and 2p@xmath1 @xmath2states .
this is in a good agreement with previously published value 2.17a.u .
obtained from the eigenchannel _
r_-matrix calculations @xcite . to understand the effect of the ir laser pulse it is a necessary prerequisite to calculate the static photoabsorption spectrum in the absence of the dressing laser pulse .
the static photoabsorption yield corresponds to a fano line - shape @xcite across the known energy @xmath118 of the 2s2p ( @xmath0)doubly - excited state with respect to the 1s@xmath1ground state , which has a known width @xmath119 and value @xmath120 of the fano line - shape parameter .
however , it is a stringent test of our basis set quality and of the cap used throughout this work to calculate this photoabsorption yield independently using the present hamiltonian and compare the corresponding spectrum with previously published results @xcite .
this can be done by setting @xmath121 in eq .
( [ eq : approxmat ] ) .
taking into account the low intensity of the xuv laser field , it is most tractable to calculate the static photoabsorption spectrum using time - dependent perturbation theory . the corresponding loss of population of the ground state
is shown in fig .
[ fig : statlos ] .
this curve should correspond to the convolution of the fano line - shape with the the xuv laser pulse bandwidth .
therefore , we have fitted the absorption yield obtained from the perturbation calculations to this convolution and extracted the position , width and line - shape parameter of the 2s2p ( @xmath0)autoionizing state .
the fitting procedure assumes a gaussian envelope of the laser pulse with bandwidth 0.06ev ( fwhm ) .
the values for position and width obtained from the fitting are compared with previously published results in table [ tab : respos ] . as expected , they agree very well with the values obtained from diagonalization of @xmath122 . in addition , the fitting procedure yields @xmath123 , in good correspondence with the value -2.75 published in ref .
@xcite . the convolved fano line - shape corresponding to the fitted parameters is also shown in fig .
[ fig : statlos ] .
the good correspondence between the perturbation calculations ( using the @xmath34 xuv pulse envelope ) and the fano line - shape fit convolved with the gaussian function suggests that the results are not sensitive to the details of the xuv pulse envelope .
the loss of the ground state population at the end of the probe pulse @xmath124 calculated from eq .
( [ eq : approxmat ] ) as a function of the xuv photon energy @xmath125 for several time delays @xmath15 is shown in fig .
[ fig : lospop ] .
nm and peak intensity @xmath126w / cm@xmath1 ] since the duration of the dressing laser pulse is longer than the duration of the xuv pulse , different intensities of the ir field are probed at different time delays @xmath15 . as can be seen in fig .
[ fig : lospop ] , the fano line - shape disappears in the case of the perfect overlap of both pulses and the photoabsorption yield becomes essentially structureless .
the peak and the minimum shifted with respect to the static case become more pronounced with increasing time delay @xmath15 .
an additional peak appears around the xuv photon energy 60.6ev for @xmath128fs .
its energy position does not change significantly with changing time delay .
@xcite performed a model calculation of the photoabsorption yield based on the three - level model developed in ref .
@xcite describing the autler - townes doublet formation .
it explains the presence of the additional peak in the photoabsorption yield as well as the shift of the higher peak with respect to the static case .
the ref .
@xcite model calculations that include optical field ionization of the doubly - excited states into the @xmath116 continuum due to the ir field confirm that this effect leads to a significant line - width broadening .
this broadening makes the autler - townes doublet nearly invisible and the position of the peak at positive xuv detunings hardly shifts as the ir pulse intensity is changed .
since the calculations presented in this work were performed at an intensity that allows for optical field ionization , our results are in good correspondence with the theoretical considerations presented in ref .
@xcite .
the present calculations account for the time profile of both pulses , whereby the photoabsorption yields shown in fig .
[ fig : lospop ] together with the static photoabsorption yield plotted in fig .
[ fig : statlos ] , enable a calculation of the transient absorption spectra that can be compared with experimental results published in @xcite .
the change of the optical density @xmath129 is calculated using the formula that takes into account the energy resolution of the spectrometer used in the experiment @xcite : @xmath130d\omega}{\int\exp\left[-\sigma_0(\omega)n_t\ell-\frac{4\ln 2}{\delta\omega^2}\left(\omega-\omega_x\right)^2\right]d\omega } , \label{eq : transab}\ ] ] where @xmath131 and @xmath132 are the photoabsorption cross section of the dressed medium and static photoabsorption cross section in the absence of the ir pulse , respectively .
@xmath133 is the atom number density of the helium gas in the tube and @xmath134 is the length of the tube in the direction parallel to the direction of laser propagation .
@xmath135ev is the energy resolution of the spectrometer ( fwhm ) used in the experiment @xcite .
determination of the photoabsorption cross section from the photoionization yield obtained from eq .
( [ eq : approxmat ] ) is not straightforward .
the usual relation between the single - photon absorption cross section and the photoabsorption rate ( probability per photon per second ) @xmath136 for an infinitely long monochromatic pulse with photon energy @xmath137 is given by the formula @xmath138 where @xmath136 is the photoabsorption rate as a function of the photon energy @xmath137 and @xmath139 is the corresponding light intensity .
this formula assumes that the photoabsorption rate does not depend on time .
the situation in the present work is more complicated .
both pulses have a finite duration , and the photoabsorption rate varies with time .
this time dependence should be included in eq .
( [ eq : csgeneral ] ) . however , when one takes into account other assumptions made regarding the shape of the xuv pulse , it is a good approximation to calculate the photoabsorption cross section using the formula @xmath140 where @xmath141 is the photoabsorption probability obtained from eq .
( [ eq : approxmat ] ) , @xmath142 is the duration of the xuv pulse and @xmath143 is the corresponding xuv pulse intensity used to calculate @xmath141 .
the scaling factor @xmath144 approximately accounts for the time dependence of the xuv field intensity and of the photoabsorption rate .
its value was chosen so that the calculated absorption spectra reproduce well the experimental transient absorption spectra published in ref .
since @xmath144 is related to the xuv pulse only , the value used to calculate the static optical density is the same as the value used to calculate the optical density in the presence of the ir field .
the comparison between our theoretical results and the ref .
@xcite experiment is presented in fig .
[ fig : expercompar ] . ) for different time delays @xmath15 and their comparison with experimental results @xcite .
the value of the scaling constant @xmath145 was used to calculate the spectra with time delays 0fs , 20fs and 40fs .
the value @xmath146 was used to calculate the spectra with time delays 60fs and 80fs . ] the best correspondence between the present calclations and experimental results was achieved by setting @xmath145 for the time delays @xmath147fs , 20fs and 40fs .
the remaining spectra with time delays @xmath148fs and 80fs compare best with the experimental results for @xmath146 .
therefore , the best fit of the experimental data is obtained when the product @xmath149 used in the calculations is smaller than the values estimated in the experimental work @xcite .
the calculated transient absorption spectra plotted in fig .
[ fig : expercompar ] have been shifted on the energy axis by 0.45ev to correct for the inaccuracy of the 2s2p ( @xmath0)doubly - excited state position with respect to the ground state in the present calculations ( see table [ tab : respos ] and the discussion in sec .
[ sec : staticfree ] ) .
a change of the value of the parameter @xmath144 affects the ratio of the peak magnitudes in the transient absorption spectra , however it does not affect the positions of the structures in the curves . as can be seen in fig .
[ fig : expercompar ] , the present theoretical calculations are generally in encouraging agreement with the experimental results .
the left peak that develops with increasing time delay is a consequence of the rising maximum of the photoabsorption spectrum due to decreasing ir intensities probed by the xuv pulse .
the figure shows that this peak develops in the present calculations more rapidly with increasing time delay than is evident in the experimental results .
the peak at positive xuv detunings calculated in the present study corresponds well to the experimental results . since the probe pulse is produced by the hhg , it is expectable that its spatial and frequency profile have more difficult structure than a simple gaussian envelope as it is assumed in this work . therefore , the discrepancies between the experimental transient absorption spectra published in ref .
@xcite and the spectra presented here can possibly be attributed to the simplified shape of the xuv pulse assumed in the present theoretical treatment .
although the photoabsorption yields in fig . [ fig : lospop ] show the reduced xuv photon absorption caused by the dressing ir laser pulse , the relation between these results and the eit - like structures is not straightforward . the spectral peaks corresponding to the dressed states of the autler - townes doublet approach each other linearly as the rabi splitting @xmath150 of the states coupled by the ir laser pulse decreases .
this behavior is not apparent in fig .
[ fig : lospop ] .
the rabi splittings @xmath150 probed by the peak of the xuv laser pulse at time delays @xmath147fs , 20fs and 40fs are @xmath151ev , 0.62ev and 0.0025ev , respectively .
the splitting for larger time delays @xmath15 is significantly smaller than the width of the 2s2p ( @xmath0)doubly - excited state .
however , fig .
[ fig : lospop ] does not show any corresponding rapid change in the energy separation of the peaks as the time delay is changed .
the clear interpretation of these photoabsorption yields is complicated by the optical field ionization of the doubly - excited states , by the finite durations of the laser pulses , and by the large detuning of the dressing laser from the energy difference of the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states . in order to eliminate the effect of the large ir detuning ,
another calculation was performed with the wavelength of the ir laser pulse modified to 600 nm .
the corresponding photon energy 2.06ev now matches the calculated energy difference between the doubly - excited states 2s2p ( @xmath0)and 2p@xmath1 @xmath2(see table [ tab : respos ] ) . in order to suppress the effect of the optical field ionization ,
the peak intensity of the coupling ir laser pulse was decreased to @xmath152w / cm@xmath1 .
corresponding photoabsorption yields are shown in fig .
[ fig : resonir - weak ] .
w / cm@xmath1 and for the dressing pulse wavelength @xmath153 nm corresponding to the energy difference between the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states .
curves for time delays @xmath147fs , 20fs 40fs are compared with the static photoabsorption yield with no ir pulse ] this graph shows for @xmath147fs two peaks with energy separation 0.2ev and minimum between them at energy close to the position of the static photoabsorption peak .
the rabi splitting of the autoionizing states coupled by the ir laser pulse with the reduced peak intensity used here is @xmath154ev .
the curve for time delay @xmath155fs shows similar behavior .
the energy separation of the peaks is 0.16ev , the rabi splitting of the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states probed by the peak of the xuv laser pulse is @xmath156ev . in case of larger time delay @xmath157fs
both peaks have a strong overlap and their energy separation is smaller than the width of the 2s2p ( @xmath0)doubly - excited state .
this correspondence confirms that we can interpret these structures as the autler - townes doublet .
energy separations of the peaks for single time delays @xmath15 do not exactly match the corresponding rabi splittings @xmath150 , mainly because of the finite duration of both pulses .
another reason could be the coupling of additional doubly - excited states by the ir laser pulse that would add further complications to the three - level picture .
the presence of an additional autoionizing state in the coupling scheme also appears to explain the smallest additional peak at an energy below 59.4ev which occurs at small time delays in fig .
[ fig : resonir - weak ] .
the 2s@xmath1 @xmath2doubly - excited state is a good candidate , since the energy difference between the 2s@xmath1 @xmath2and 2s2p ( @xmath0)doubly - excited states is 2.373ev in the present calculations . in general , the calculations performed with the reduced ir intensity and with the photon energy adjusted as discussed above show the eit - like structure more clearly than the calculations performed with parameters of the ir laser pulse used in the experiment .
this suggests that the autler - townes structure in the calculations carried out using the experimental parameters are not very pronounced simply because the dressing laser detuning is so large . in order to investigate the manner in which the optical field ionization of the autoionizing states affects the dressed photoabsorption spectra , another calculation with the adjusted ir laser pulse ( @xmath153 nm )
was performed .
the experimental peak intensity of the ir dressing laser pulse @xmath126w / cm@xmath1 was used .
the resulting photoabsorption yields for several time delays @xmath15 are shown in fig .
[ fig : resonir - strong ] .
w / cm@xmath1 and for the dressing pulse wavelength @xmath153 nm corresponding to the energy difference between the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states .
curves for time delays @xmath147fs , 20fs 40fs are compared with the static photoabsorption yield . ]
this graph shows a broad peak around the energy 58.7ev and a richer structure in the energy interval from 59.7ev to 60.3ev in case of perfect temporal overlap of the two laser pulses .
these structures approach each other with increasing time delay @xmath15 .
since optical field ionization broadens the line - widths , the characterization of the individual peaks in the photoabsorption spectra is not straightforward .
however , the well pronounced separation of the broad peak at negative xuv detuning and the structure at positive xuv detuning is most likely related to the coupling of the 2s2p ( @xmath0)and 2p@xmath1 @xmath2autoionizing states , as can be assumed from the change of the separation with increasing time delay @xmath15 .
it is possible that the ir field at the experimental peak intensity @xmath126w / cm@xmath1 enables coupling of additional autoionizing states and this coupling leads to the rich structure of peaks appearing at positive xuv detunings .
this work presents theoretical treatment of the photoabsorption in xuv for helium dressed by the ir laser .
the 2s2p ( @xmath0)and 2p@xmath1 @xmath2doubly - excited states are coupled by strong 800 nm laser pulse with duration 42fs .
this system is probed by the weak xuv pulse produced by hhg .
the photoabsorption yields in the presence of both pulses are calculated as a function of the xuv photon energy for several dressing - probe time delays by solving the tdse .
the calculated transient absorption spectra are in reasonable correspondence with experimental results of @xcite .
the presence of the coupling ir laser significantly reduces the peak photoabsorption yield and changes of the time delay between the dressing and probe pulse show the dependence on the ir field intensity that is probed by the xuv photons .
calculations performed with the wavelength of the dressing laser modified to @xmath153 nm , corresponding to zero detuning of the ir laser , clearly show a structure in the photoabsorption spectrum that can be interpreted as the autler - townes doublet and an eit - like structure .
this allows us to conclude that this was not clearly pronounced in the experimental results published in ref .
@xcite primarily because of the large detuning of the ir laser .
the photoabsorption yields calculated for different dressing laser intensities showed that the optical field ionization of the doubly - excited states broadens the linewidths and increasing dressing laser intensity leads to formation of more complicated structures in the photoabsorption spectra .
these could result from the increasing importance of additional autoionizing states at higher ir intensity .
the method of solution of the tdse presented in this work proved to be stable and suitable for theoretical treatment of two active electron atoms in pump - probe settings of the laser pulses with a potential of further improvement in the future .
conversations with zhi - heng loh and stephen leone are greatly appreciated , as is access to some of their unpublished data and calculations .
we also thank andreas becker and antonio picn for stimulating discussions . this work was supported in part by the department of energy ,
office of science .
this research used resources of the national energy research scientific computing center , which is supported by the office of science of the u.s .
department of energy .
since it is necessary to integrate the system of equations ( [ eq : tdmat ] ) for a set of different xuv photon energies @xmath14 and pump - probe time delays @xmath15 , it is important to employ an integration method that efficiently meets these requirements . after tests of several different methods used elsewhere in similar contexts ( e.g. runge - kutta @xcite and crank - nicholson @xcite ) we chose to implement the split - operator technique @xcite in a basis set representation .
the basic idea of this propagation method is that if the total time - dependent hamiltonian @xmath158 can be written as a sum of the time - independent and time - dependent parts @xmath159 , the propagation of the wave function between times @xmath160 and @xmath161 , where @xmath162 is sufficiently small , can be implemented to second order in the time step , as follows : @xmath163\exp\left(-\frac{i\delta t}{2}\mathbf{h}_0\right)\mathbf{\psi}(t)+\mathbf{o}\left[\left(\delta t\right)^3\right ] .
\label{eq : splitop}\ ] ] if we denote @xmath164 then the application of this method to the present system of equations ( [ eq : approxmat ] ) yields @xmath165 \exp\left[-i\mathbf{d}_z^if_i(t+t_d,\delta t)-i\mathbf{d}_z^xf_x(t,\delta t)\right]\nonumber\\ \exp\left[-\frac{i\delta t}{2}\left(\mathbf{h}+\mathbf{v}_c\right)\right]\mathbf{c}(t)+\mathbf{o}\left[\left(\delta t\right)^3\right ] .
\label{eq : basprop } \end{aligned}\ ] ] the first and third exponential terms in this equation can be calculated rather easily , as they are time independent ( for a fixed time step @xmath162 ) and all the matrices are block diagonal .
since the eigenrepresentation of the field - free hamiltonian @xmath44 is used , this term can be expressed using the crank - nicholson formula @xmath166=\frac{\mathbf{1}-i\left(\mathbf{h}+\mathbf{v}_c\right)\delta t/4}{\mathbf{1}+i\left(\mathbf{h}+\mathbf{v}_c\right)\delta t/4}+\mathbf{o}\left[\left(\delta t\right)^3\right ] , \label{eq : cnpart}\ ] ] which is accurate to second order in the time step .
although the evaluation of this term requires a matrix inversion , it is not demanding , as it can be performed independently for every block matrix corresponding to a definite angular momentum @xmath47 .
in addition , it is necessary to perform this step only once before calculating the time propagation , as this term is time independent .
calculation of the second exponential in eq .
( [ eq : basprop ] ) is more difficult , as it explicitly depends on time and is not block diagonal . since @xmath82 and @xmath83 do not commute ( see the structure in fig .
[ fig : eqstr ] ) , we can write this term using the split - operator formula ( [ eq : splitop ] ) as follows : @xmath167=\exp\left[-\frac{i}{2}\mathbf{d}_z^if_i(t+t_d,\delta t)\right]\nonumber\\
\exp\left[-i\mathbf{d}_z^xf_x(t,\delta t)\right]\exp\left[-\frac{i}{2}\mathbf{d}_z^if_i(t+t_d,\delta t)\right]+\mathbf{o}\left[\left(\delta t\right)^3\right ] .
\label{eq : secondsplit } \end{aligned}\ ] ] the most straightforward way to calculate these exponentials is using the eigenrepresentations of @xmath82 and @xmath83 : @xmath167=\mathbf{u}_i\mathrm{diag}\left\{\exp\left[-\frac{i\lambda^i_k}{2}f_i(t+t_d,\delta
t)\right]\right\}\mathbf{u}_i^t\nonumber\\ \mathbf{u}_x\mathrm{diag}\left\{\exp\left[-i\lambda^x_kf_x(t,\delta t)\right]\right\}\mathbf{u}_x^t\mathbf{u}_i\mathrm{diag}\left\{\exp\left[-\frac{i\lambda^i_k}{2}f_i(t+t_d,\delta t)\right]\right\}\mathbf{u}_i^t , \end{aligned}\ ] ] where @xmath168 are the eigenvalues of @xmath169 and @xmath170 are the corresponding matrices of the column eigenvectors .
substitution of this result together with eq .
( [ eq : cnpart ] ) into eq .
( [ eq : basprop ] ) and denoting @xmath171 yields @xmath172\right\}\mathbf{v}_2\mathrm{diag}\left\{\exp\left[-i\lambda^x_kf_x(t,\delta t)\right]\right\}\mathbf{v}_2^t\nonumber\\ \mathrm{diag}\left\{\exp\left[-\frac{i\lambda^i_k}{2}f_i(t+t_d,\delta t)\right]\right\}\mathbf{v}_1^t\mathbf{c}_i(t).\qquad \label{eq : irprop } \end{aligned}\ ] ] this propagation scheme in one time step means that the wave function is first expressed in the eigenrepresentation of @xmath82 and propagation for a quarter of the time step is performed using the field - free hamiltonian augmented by the cap .
then the wave function is propagated in the ir field for half of the time step .
after that the wave function is transformed into the eigenrepresentation of the xuv dipole operator and the propagation for one time step is performed in the xuv field only .
the wave function is transformed back to the ir eigenrepresentation again and the propagation steps in the ir field only as well as the step using the field - free hamiltonian are performed .
this propagation scheme involves only two matrices which are time dependent and they are diagonal in different representations .
it means that only two changes of representation are necessary in every time step .
note that application of eq .
( [ eq : irprop ] ) in two subsequent time steps allows for multiplication by the matrix @xmath173 between the steps .
therefore , every propagation step involves three matrix - vector multiplications and two diagonal matrix - vector multiplications , where only the diagonal matrices are time dependent .
the matrix @xmath174 is real and the matrix @xmath175 is complex symmetric . after the transformation of the initial state vector @xmath176 from the energy eigenrepresentation of @xmath71 to the diagonal representation of @xmath82 , the rest of the propagation is performed in this basis .
this approach is permissible in the context of the problem studied here , because we only need to determine the depletion of the field - free ground state at the end of the xuv laser pulse only , rather than information about the evolution of the complete wave function during the pulse .
note that only the function @xmath177 depends on the xuv photon energy @xmath14 and the information about the time delay @xmath15 appears only in the diagonal matrix in eq .
( [ eq : irprop ] ) .
therefore , the same matrices @xmath173 and @xmath174 can be used in all the time dependent calculations for all the values of @xmath178 and @xmath15 of interest .
one more note related to eq .
( [ eq : secondsplit ] ) should be made at this point .
the split operator method is used there to calculate the exponential involving matrices @xmath82 and @xmath83 which do not commute .
the non - zero commutator is a consequence of the approximation introduced in sec .
[ sec : theo ] , where we neglect the effect of the ir laser field on the ground state and neglect the xuv couplings that do not involve the ground state ( see fig .
[ fig : eqstr ] ) .
validity of this approximation implies that the commutator @xmath179 $ ] is sufficiently small to write the exponential on the left hand side of eq .
( [ eq : secondsplit ] ) simply as a product of two exponentials .
this would change the form of the final time propagator ( [ eq : irprop ] ) in such way that it will contain only two time - dependent matrices .
therefore , this approach would save one matrix - vector multiplication in every time step .
however , its impact on the accuracy of the time propagation needs to be tested and its implementation will be the subject of forthcoming research . in conclusion ,
the time required to calculate the matrices @xmath173 and @xmath174 ( matrix diagonalization , inversion and the matrix - matrix multiplication ) is negligible compared to the calculations of the time propagation using eq .
( [ eq : irprop ] ) .
this propagation scheme is of the second order in the time step and scales as @xmath180 with the dimension of the basis set used ( since it involves only matrix - vector multiplications in every time step ) .
31ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop _
_ ( , ) link:\doibase 10.1103/physrevlett.64.1107 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.77.633 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.75.457 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.94.023002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.143002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.053002 [ * * , ( ) ] link:\doibase 10.1103/physreva.34.4785 [ * * , ( ) ] http://stacks.iop.org/0953-4075/37/i=21/a=005 [ * * , ( ) ] link:\doibase 10.1103/physrev.124.1866 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.98.253001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.193008 [ * * , ( ) ] link:\doibase 10.1103/physreva.81.063403 [ * * , ( ) ] link:\doibase 10.1103/physreva.62.062719 [ * * , ( ) ] link:\doibase 10.1103/physreva.53.1424 [ * * , ( ) ] http://stacks.iop.org/0953-4075/28/i=15/a=010 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physreva.69.032707 [ * * , ( ) ] link:\doibase 10.1103/physreva.77.032716 [ * * , ( ) ] link:\doibase 10.1016/s0370 - 1573(98)00002 - 7 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1103/physreva.59.r1742 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.101.233001 [ * * , ( ) ] http://stacks.iop.org/0953-4075/38/i=22/a=005 [ * * , ( ) ] http://books.google.co.in/books?id=99k9aqaaiaaj[__ ] ( , ) link:\doibase 10.1103/physreva.76.043420 [ * * , ( ) ] @noop _ _ ( , , ) | electromagnetically induced transparency - like behavior in the extreme ultraviolet ( xuv ) is studied theoretically , including the effect of intense 800 nm laser dressing of he 2s2p ( @xmath0)(@xmath0 ) and 2p@xmath1 @xmath2(@xmath3 ) autoionizing states .
we present an _ ab initio_solution of the time - dependent schrdinger equation ( tdse ) in an _ ls_-coupling configuration interaction basis set .
the method enables a rigorous treatment of optical field ionization of these coupled autoionizing states into the @xmath4 continuum in addition to @xmath5 .
our calculated transient absorption spectra show encouraging agreement with experiment . |
global and local , private and public institutions increasingly rely on numerical indexes for their decision making processes
. a single index may be the difference between approval or rejection .
these indexes , and even the mathematical models behind them , are often made public to enhance transparency , accountability , and establish standards .
the usefulness of an index is measured by the successes and failures of decisions based on it .
thus index creation , computation , selection , comparison , evaluation , and consolidation has become an industry of its own , and an integral part of institutionalized decision making processes .
+ in this work we address the problem of finding suitable indexes for ranking countries according to their influences on the international trade market . in this case
there is an obvious candidate : the influence of a country on the international trade market is proportional to the amount ( counted in us dollars ) of its international trade , i.e. the total amount of imports plus the total amount of exports .
this primordial index is sound and should not be overlooked .
nevertheless , we claim that it disregards some subtle but important issues .
+ suppose we have a couple of countries @xmath0 and @xmath1 both with high levels of international trade , so that they both are highly ranked with the above index .
suppose in addition that @xmath0 and @xmath1 trade essentially with each other , i.e. trade of @xmath0 and @xmath1 with other countries is negligible in comparison with trade between themselves . in this case
our feeling is that @xmath0 and @xmath1 do not exert a strong influence on the international trade market : a disruption on @xmath0 s economy will surely impact @xmath1 s economy , but will have a negligible impact on global trade .
countries @xmath0 and @xmath1 although highly interdependent are in fact quite isolated from the rest of the world , and should not be ranked as highly influential countries in the international trade market .
+ to address this sort of issues we take a network approach @xcite to our index creation problem . in section [ stn ] , we regard the international trade market as a weighted network with nodes representing countries , edges representing trade between countries , and weights measuring the influence that a country exerts on another country trough trade . indeed , in sections [ dit ] and [ dio ] , we are going to introduce a couple of different weights leading to a couple of rankings .
+ once we have established the network settings for approaching our index creation problem , we face the problem of ranking nodes in a network by their influence .
this demands that we paid attention to the distinction between direct and indirect influences in a network , a distinction emphasized by godet and his collaborators @xcite , who stressed the power of studying indirect influences for uncovering hidden relations .
direct influences in a network arise from directed edges .
indirect influences in a network arise from chains of direct influences , i.e. from directed paths .
+ let us consider an extreme case that illustrates the importance of taking indirect influences into account .
cuba and the united states ( us ) have a fairly weak amount of trade , indeed since 1960 the us has placed a series of economic , financial , and commercial prohibitions to trade with cuba .
thus a fairly low influence of the us on the cuban economy is to be expected .
however , the us trades with cuba s main trading partners , and so we may expect that the us exerts a stronger indirect influence on the cuban economy than what one may naively think
. simply put , a disruption of the us s economy will likely impact the spaniard economy , and trough spain it will also be felt by the cuban economy .
the main challenge that we confront in this work is to give a quantitative account of the latter impact . +
after building our mathematical models in sections [ stn ] trough [ hiitn ] , we proceed in sections [ secactn ] and [ wtn2 ] to implement them using real world data . out of united nations
member states we restrict our attention to 177 countries , namely , those that have recent data available in the economic commission for latin america and the caribbean ( eclac ) web site @xcite .
we collect the 2011 data of exports and imports between each pair of countries , as well as the gdp , and the exports and imports totals for each country . using this data
we build the `` international trade network , '' where nodes are countries , and a pair of nodes is connected by an edge if there were exports or imports between them in 2011 .
suppose there actually was trade between countries @xmath0 and @xmath1 , then we define a couple of weights on the edge from @xmath0 to @xmath1 as follows : 1 . * direct influences on trade .
* this weight computes the proportion of the international trade of @xmath1 ( exports + imports ) that involves country @xmath2 + 2 . * direct influences on offer .
* this weight computes the relative contribution of the trade between @xmath0 and @xmath1 to the offer of @xmath1 ( gdp + imports ) . to take indirect influences into account we start from a trade network , with one of the weights just introduced , and apply one of the mathematical methods available for computing indirect influences in complex networks . with any of these methods
one goes beyond computing direct trade between countries @xmath0 and @xmath1 , and takes as well into account trade chains @xmath3 such that country @xmath4 trades with country @xmath5 .
+ in this work we weight chains of trade using the pwp method introduced by daz @xcite . to place this method whiting a general context and for the reader convenience we provide a in section [ mcii ] a succinct description of four closely related methods for computing indirect influences , namely , godet s micmac , google s pagerank , chung s heat kernel , and the pwp method . for more details on the similarities and differences among these methods
the reader may consult @xcite .
our choice of the pwp method rest on the fact that with it any chain of direct influences , of any length , generates indirect influences , and reciprocally , all indirect influences are generated in this fashion .
+ thus we obtain several rankings among nations using the methodology outlined above : for each of the above weights we get the direct and indirect influences rankings .
we compare these rankings among themselves and with the gdp ranking for the american continent trade network in section [ secactn ] , and for the world trade network in section [ wtn2 ] .
we show that there are remarkable differences between the various rankings , and analyse them in economic terms .
+ we remark once again that our economic data refers to year 2011 , and is measured in us dollars . when we compare different indexes what is really at stake is to compare the rankings they induce . in practice what we do is to compare the corresponding normalized indexes .
numerical calculations in this work were made with scilab s module for computing indirect influences designed by catumba @xcite .
for the economics terminology the reader may consult @xcite .
in this section we introduce basic definitions concerning the construction of international trade networks and their adjacency matrices , called matrices of direct influences in this work . a trade network @xmath6 is a directed graph such that : * the set @xmath7 of vertices is a family of countries .
* there is an edge in @xmath8 from vertex @xmath0 to vertex @xmath1 if and only if there is trade ( exports or imports ) between @xmath0 and @xmath1 . note that trade networks are actually symmetric graphs
nevertheless , we regard them as directed graphs since we are soon going to introduced non - symmetric weights on them . for each edge
@xmath9 , let @xmath10 be its source vertex , or equivalently , the country that exerts the influence , and let @xmath11 be its target vertex , or equivalently , the country that is influenced .
a weight @xmath12 on a directed graph @xmath13 assigns weight @xmath14 to each edge of @xmath13 , i.e. a weight on @xmath13 is a map @xmath15 . in sections [ dit ] and [ dio ] we are going to introduce a couple of weights on trade networks .
+ the bi - degree @xmath16 of a vertex @xmath17 in a weighted directed graph @xmath13 is such that @xmath18 is the sum of the weights of edges with target @xmath0 , and @xmath19 is the sum of the weights of edges with source @xmath0 , that is : @xmath20 we impose the alphabetic order on countries , so the set of countries @xmath7 is identified with the set @xmath21 , where @xmath22 is the number of countries in our trade network .
the adjacency matrix or matrix of direct influences @xmath23 of a trade network @xmath13 is given for @xmath24 by : @xmath25 thus @xmath26 gives the direct dependence of @xmath0 on @xmath1 , or equivalently , the direct influence of @xmath1 on @xmath0 .
the matrix of indirect influences @xmath27 is computed from the matrix of direct influences @xmath23 using one of the methods discussed in section [ mcii ] . in our applications , the matrix of indirect influences @xmath28 is computed applying the pwp method . [ d1 ]
let @xmath13 be a trade graph , @xmath23 its matrix of direct influences , and @xmath27 its associated matrix of indirect influences .
the indirect dependence @xmath29 , indirect influence @xmath30 , and indirect connectedness @xmath31 of vertex @xmath0 in @xmath13 are given , respectively , by : @xmath32 the ordered pair @xmath33 is the bi - degree of @xmath0 in the network of indirect influences .
the indirect connectedness of vertices @xmath0 and @xmath1 is given by @xmath34 direct dependencies , influences , and connectedness are computed in a similar way using the matrix of direct influences @xmath23 instead of the matrix @xmath28 of indirect influences . following godet a @xmath35-dimensional representation of vertices bi - degrees can be displayed trough the dependence - influence plane which comes naturally divided in four sectors , see figure [ p1 ] .
the horizontal axis represents dependencies and the vertical axis represents influences .
a country @xmath0 is represented in the dependence - influence plane by the ordered pair @xmath33 .
the horizontal and vertical lines defining the four sectors of the plane are located at the mean dependence @xmath36 and at the mean influence @xmath37 , given respectively , by @xmath38 * * sector 1 : * influential independent countries . * * sector 2 : * influential dependent countries . * * sector 3 : * low influence independent countries . * * sector 4 : * low influence dependent countries .
although our focus is on the pwp method , for the reader convenience we introduce four methods for computing indirect influences , and describe how each method computes the matrix of indirect influences . for more on the similarities and differences among these methods
the reader may consult @xcite . with this method , introduced in 1992 by godet @xcite ,
the matrix of indirect influences is @xmath39 where @xmath23 is the matrix of direct influences , and @xmath40 is a parameter usually equal to @xmath41 or @xmath42 .
the relevant paths in the network , with this method , are those of length @xmath43 .
this method , registered in 1999 by google @xcite , is quite well - known thanks to its application to web searching . with pagerank influences
are normalized , and dependencies measure the relevance or relative importance of web pages .
pagerank takes infinite powers of matrices , thus it gives greater importance to infinite long paths .
let @xmath23 be the matrix of direct influences , whose entries should be non negative real numbers , and the sum of each column must be either @xmath44 or @xmath45 .
the matrix of indirect influences is given by @xmath46^{k}$ ] , where @xmath47 is the matrix obtained form @xmath23 by replacing the entries of a zero column by @xmath48 , @xmath49 is the matrix with entries @xmath48 , and the parameter @xmath50 is usually @xmath51 .
+ in the web network vertices are web pages , and there is an edge from page @xmath1 to page @xmath0 if there is a hyperlink in @xmath1 leading to @xmath0 .
the matrix of direct influences @xmath23 is given by : @xmath52 where @xmath53 is out - degree of the vertex @xmath1 .
+ the ranking of page @xmath0 is proportional to its indirect dependence @xmath54 .
a higher rank means an earlier showing in a web search .
this method was introduced in 2007 by chung @xcite .
let @xmath23 be the matrix of direct influences , then the matrix of indirect influences is given by @xmath55 where @xmath56 is the identity matrix with the size of @xmath23 , and @xmath57 is a fixed positive real number .
this method was introduced in 2009 by daz @xcite . with pwp , as well as with micmac and heat kernel
, one can compute indirect influences for any matrix of direct influences , even one with negative entries . fix a positive real number @xmath57 ( set to @xmath45 in our applications ) and let @xmath23 be the matrix of direct influences .
the matrix of indirect influences is given by : @xmath58 where for a matrix or a number @xmath59 we set @xmath60
we go back to the problem of defining direct influences on a trade network @xmath13 .
[ ddit ] the direct influence on trade @xmath61 of country @xmath1 on country @xmath0 is given by : @xmath62 where : * @xmath63 is the amount of exports from @xmath1 to @xmath0 , according to @xmath0 s data .
* @xmath64 is the amount of imports from @xmath1 to @xmath0 , according to @xmath0 s data . * @xmath65 and @xmath66 are , respectively , the total exports and total imports of @xmath0 .
the direct influence on trade @xmath61 measures the portion of @xmath0 s international trade that involves @xmath1 : out of each dollar the @xmath0 trades with the world , it trades @xmath61 dollars with @xmath1 .
a few remarks on the definition of direct influences on trade @xmath61 : + * exports from country @xmath0 to country @xmath1 are often though as creating dependence of @xmath1 on @xmath0 . in our approach exports
creates dependence in both directions : @xmath1 needs some products coming from @xmath0 ( otherwise it would not buy ) , just as @xmath0 needs the currency coming from @xmath1 ( otherwise it would not sell . ) + * dependencies are normalized , indeed we have that @xmath67 * a country trading significatively with many different partners will , in general , have low dependence on any particular country .
us and china are the first and the second world largest economies , respectively .
thus their mutual influences are a topic of global interest , see figure [ hoy ] in which the diameter of vertices is roughly proportional to gdp .
definition [ ddit ] gives us a quantitative method for measuring these influences .
+ according to us s data its exports and its imports in 2011 ( in thousands of us dollars ) were 1,479,730,169 @xmath68 2,262,585,634 , respectively ; while its exports to and its imports from china were 103,878,414 @xmath68 417,302,859 , respectively .
so the influence of china on the us s international trade is : @xmath69 thus a @xmath70 of the us international trade involves china ( either as a buyer or as a seller . )
+ according to china s data its exports and imports in 2011 were 1,898,388,435 @xmath68 1,743,394,866 , respectively ; while its exports to and its imports from the us were 325,010,987 @xmath68 123,124,009 , respectively .
so the influence of the us on china s international trade is : @xmath71 thus a @xmath72 of china s international trade involves the us .
+ so , with respect to trade , china has a higher direct influence on the us than viceversa , and influences in both directions are quite high .
the definition of direct influences given in the previous section measures international trade . in this section
we take a different approach that considers both international trade ( trough exports and imports ) and domestic trade ( trough gdp ) .
[ ddio ] the direct influence @xmath26 of country @xmath1 on the total offer of country @xmath0 is given by : @xmath73 where : * @xmath63 is the amount of exports from @xmath1 to @xmath0 , according to @xmath0 s data .
* @xmath64 is the amount of imports from @xmath1 to @xmath0 , according to @xmath0 s data . *
gdp@xmath74 and @xmath66 are @xmath0 s gross domestic product and imports total , respectively .
the direct influence on offer @xmath26 measures the relative size of the trade between @xmath0 and @xmath1 in comparison with @xmath0 s total offer gdp@xmath75 .
in other words , @xmath26 is the portion of products trade in @xmath0 ( both home grown and imported ) that involve @xmath1 ( either because they were imported from or exported to @xmath1 ) .
+ a couple remarks on the definition of direct influences on offer @xmath26 : + * the total dependence @xmath54 of country @xmath0 measures the portion of its offer that involves ( trough imports or exports ) the rest of the world .
indeed we have that @xmath76 * the greater the gdp of a country , the lesser its dependence on other countries .
let us compute the china - us bilateral influences on offer .
the gdp of china and us amounted in 2011 to 7,321,935,025 @xmath68 14,991,300,000 , respectively .
thus the influence of china on the us s offer is given by : @xmath77 thus a @xmath78 of the us s offer involves china ( either as a buyer or as a seller . )
+ the influence of the us on china s offer is in turn given by : @xmath79 thus a @xmath80 of china s offer involves the us .
+ so , regarding offer , we find that the us have a higher influence on china than viceversa , reversing the result obtained for direct influences on trade in section [ dit ] .
to account for indirect influences on a trade network we apply the pwp method from section [ mcii ] to the matrices @xmath81 and @xmath23 of definitions [ ddit ] and [ ddio ] measuring direct influences on trade and offer .
thus we obtain the matrices indirect influences on trade and offer given , respectively , by @xmath82 let us consider in details an extreme case that illustrates the importance of taking indirect influences into account .
we use data from the world trade network to be discussed in section [ wtn2 ] . + for political reasons cuba and the us have fairly low direct trade , and as a result the direct influence of the us on cuba in trade and offer are just @xmath83 and @xmath84 , respectively , i.e. us is directly involved in @xmath85 and @xmath86 of the cuban international trade and offer , respectively . comparing direct and indirect influences of the us on cuba one obtains , after normalization , increments of @xmath87 and @xmath88 on trade and offer , respectively .
these increments in indirect vs direct influences are quite remarkable since , as we shall see in section [ wtn2 ] , the us has decreasing global indirect influences both on trade and on offer . +
the increments in indirect influences come from the fact that , beyond direct trade , the us also influences cuba trough chains of direct trade , which in the simplest case take the form of triangulations .
for example , in figure [ e1 ] we show the triangle of direct influences us - spain - cuba : the us exerts a direct influence on spain in trade and offer of @xmath89 and @xmath90 , respectively ; in turn , spain exerts a direct influence on cuba in trade and offer of @xmath91 and @xmath92 , respectively .
thus it is quite clear that the us economy exerts an indirect influence on the cuban economy trough spain .
+ computing indirect influences gives us a method for quantifying chains of direct influences . in the next sections we compute indirect influences on trade networks for a couple of scenarios
: first we consider the american continent with 35 countries , and then we consider the world wide case with 177 countries . in both cases
we compute direct and indirect influences , dependencies , and connectedness starting both from the weight based on trade , and the weight based on offer .
this generates various indexes for gauging influences in international trade .
we make a comparative analysis of the different rankings when applied to these scenarios .
we consider 35 countries located on the american continent ranging from mighty us to tiny dominica .
the trade network for the american continent is a nearly complete graph with each country trading in average with 32.2 countries out of 34 possibilities .
the fully connected countries are argentina , brazil , canada , colombia , costa rica , cuba , dominican republic , ecuador , guatemala , jamaica , mexico , peru , salvador , and the us .
the lesser connected countries are bolivia , dominica , haiti , and paraguay , trading respectively with @xmath93 and @xmath94 countries .
+ the american continent trade network , shown in figure [ casa ] , was drawn with sage using the matrix @xmath26 of direct influences on offer from definition [ ddio ] .
+ [ atnit ] the first thing one notices from the table in figure [ g2 ] is that the us , as expected , leads both in direct and indirect influences .
one , however , may wonder why its direct influences are higher than its indirect influences ?
this happens , we believe , for a couple of reasons : 1 ) because of its leading role , the us is more likely to trade with countries otherwise isolated with little influence on other countries .
2 ) since influences are normalized , the decreasing indirect influences of the us may also be explained by the opposite behaviour in countries such as canada and mexico , whose indirect influences increase to a good extend precisely because of their special relation ( nafta agreement ) with the leading country .
+ note that brazil has higher direct influences than both canada and mexico , this is to be expected since brazil s gdp is only second to the us s in the american continent . nevertheless , brazil has lesser indirect influences than both canada and mexico .
this is due to the fact that in comparison brazil is an inward looking economy , and does not have a privileged relation with the leading country .
+ the case of chile is interesting as well .
its direct influences are quite modest , since chile has a relatively small economy .
however , it has a noticeable increment in indirect influences , due to the fact that chile has an export - oriented economy .
+ trinidad and tobago ( tt ) is perhaps the most revealing case .
its small size economy seems hardly compatible with its high direct influence , and to make this case even more intriguing its indirect influences decrease quite sharply .
this curious situation may be explained as follows .
our indices measure influences on countries , and tt exerts a direct influence on quite a few low gdp countries , mostly caribbean community ( caricom ) member states .
since caricom s countries have small economies the influence of tt on them may be quite high , even if tt has itself a relatively small economy . on the other hand , many caricom member states have low direct influences , so they seldom contribute to the indirect influences of tt . in summary , the difference between direct and indirect influences of tt attests to both the leadership of tt within caricom , and the meager impact of the subregion in the economy of the american continent at large .
+ considering influences on trade in the american continent , see figure [ g1 ] , we find even less correlation between direct and indirect influences .
canada and mexico indirect influences increase quite sharply , while for tt there is a deep fall instead .
antigua goes from position @xmath95 in direct influences , to position @xmath96 in indirect influences .
bolivia has the opposite behaviour going from position @xmath97 in direct influences to position @xmath98 in indirect influences .
the world trade network containing 177 countries distributed all around the globe , shown schematically in figure [ casawtn ] , was drawn feeding sage with the matrix of direct influences on offer from definition [ ddio ] .
+ unsurprisingly , the world trade network contains a big highly connected component containing the vast majority of countries in the world , with the exception of a handful of commercially isolated countries such as buthan , guinea bissau , kiribati , samoa , tonga , and tuvalu . on average
a country trades with @xmath99 countries , and there are @xmath100 countries each trading with at least @xmath101 countries .
+ we proceed to address a couple of questions concerning the world trade network : are gdp and direct or indirect influences correlated ?
are there significative differences between direct and influences ?
these question are both analyzed by looking first to influences on offer , and then influences on trade . instead of a statistical analysis
we simply highlight interesting cases regarding the top 20 countries in indirect influences .
+ naturally , we expect high gdp countries to have high direct and indirect influences on offer , and indeed there is a generic tendency for this correlation to hold , however the actual data reveals a far more subtle situation , see figure [ casag2 ] .
+ japan has the @xmath102rd largest world gdp , but occupies the @xmath103th and @xmath104th places in direct and indirect influences , respectively .
canada is in the @xmath95th place in the gdp ranking , but occupies the 20th and 21st positions in the direct and indirect influences rankings , respectively .
similarly , mexico is in the 14th place in the gdp ranking , but 34th and 31st positions in direct and indirect influences .
the drop in influences compared to gdp in the latter cases is probably consequence of canada s and mexico s strong economic ties to the us , so their influences are somewhat diluted in the us high gdp .
+ comparing direct and indirect influences on offer , see figure [ casag2 ] , we find countries with similar results in both rankings , for example brazil , italy , switzerland , and korea .
there are countries with diminishing indirect influences as australia , china , india , russia , singapore , south africa , thailand , united arab emirates and the us . in the case of south africa and
even more so for thailand the decreases are quite sharp .
there are also countries with raising indirect influences as germany , france , japan , united kingdom , netherlands , belgium and spain . as a rule
european countries tend to have increasing indirect influences , the case of germany being remarkable because a sharp increase occurs .
+ next we consider direct and indirect influences on trade in the world trade network , see figure [ casagra ] . in this case
we expect less correlation with the gdp ranking , since we are looking at influences on international trade and disregarding domestic trade .
for example , netherlands climbs from position @xmath105 in gdp to position @xmath104 in indirect influences on trade .
similarly , belgium occupies the @xmath100th position in gdp , but raises to the @xmath106th position in indirect influences on trade . in the opposite trend , brazil is in the @xmath107th place in gdp , but is in the @xmath105th position in indirect influences on trade .
+ from figures [ casag2 ] and [ casagra ] , it is quite clear that influences on trade measure a different thing that influences on offer , even if the us , china , germany and france , in this order , remain the top 4 countries in both rankings .
for example , south africa , thailand , and the united arab emirates are not longer in the top 20 countries in influences on trade , and are replaced by canada , mexico and poland .
observe that brazil , japan , korea , netherlands , spain , and the united kingdom climb in their ranking positions , belgium remains stable , and australia , india , italy , russia , singapore , and switzerland fall in their rankings positions .
+ similarly , direct and indirect influences on trade are weaklier correlated than direct and indirect influences on trade .
only france , among the top 20 countries , have nearly identical direct and indirect influences on trade .
australia , brazil , india , italy , russia , singapore , spain , switzerland , and the us have diminishing indirect influences , whereas belgium , canada , china , germany , japan , korea , mexico , netherlands , and the united kingdom have increasing indirect influences .
the cases of germany and korea are interesting because of their sharp increments .
+ below we show the indirect dependence - influence planes , both for trade and offer , of the world trade network .
note that the united states and china have high influence and low dependence on offer , while they have both high influence and dependence on trade , this is due to their high gdps .
in contrast , germany have high influence and high dependence both on offer and on trade .
in this work we developed a network viewpoint leading to a couple of ways of ranking direct influences of countries in the international trade market , namely , a ranking based on trade and another ranking based on offer .
the first ranking measures relative international trade trough exports and imports , while the second ranking also takes into account domestic trade trough gdps . +
as far as we can tell a network approach to international trade taking into account indirect influences has not been designed so far .
most studies found in the literature gauge direct influences as evidenced by trade accounts . in this regard , they compare to our networks of direct influences in trade , rather than to our networks of indirect influences in trade . we showed that taking in consideration indirect influences may lead to more realistic indexes , as relevant cases suggest . + we computed indirect influences using the pwp method , and applied it to the american continent trade network and to the world trade network .
thus we obtained new rankings and the corresponding dependence - influence planes .
+ part of the motivation for this work is the high public interest in ranking countries for their centrality in international trade .
various rankings have been introduced , some quite sophisticated such as the dhl global connectedness index @xcite .
this ranking however results from the integration of many features , whereas our rankings focus on a single feature enhanced by taking into account indirect influences . + the problem of further integrating the various rankings proposed in the literature may be attacked by thinking of the set of rankings as a finite metric space .
for example , we can define a suitable normalized euclidean distance between indices @xmath108 $ ] ( with @xmath109 and @xmath110 being the ranking positions of country @xmath0 according to @xmath111 and @xmath112 , respectively ; and @xmath22 is the total number of countries ) given by @xmath113 furthermore , we also have the problem of consolidating the various indexes into just one index .
+ we applied our methods for computing indirect influences to trade networks at the level of countries .
it should however be clear that these methods may also be applied at the business level , and even at the individual level . in the latter cases ,
finding reliable data at a global scale and managing such huge data are daunting problems .
nevertheless , our methods can be readily applied if one focusses on specific business sectors , just like we made a focus study of indirect influences on trade and offer in the american continent .
+ the models presented in this work were static , in the sense that our data referred to just one year , namely 2011 .
nevertheless , our techniques may be extended to a multiple years dynamical model in which influences are time depended functions . | we address the problem of gauging the influence exerted by a given country on the global trade market from the viewpoint of complex networks .
in particular , we apply the pwp method for computing indirect influences on the world trade network .
[ thm]lema [ thm]definition [ thm]remark [ thm]theorem [ thm]example [ thm]corollary [ thm]proposition [ thm]table |
star clusters are subjected to various disruptive effects , which prevent low mass star clusters ( @xmath24 ) from surviving for a hubble time . from observations this was first noted by @xcite and later by , who showed that there is a lack of open star clusters older than a few gyr in the solar neighbourhood .
later , more quantitative results were obtained for the life time of star clusters ( e.g. @xcite and @xcite for clusters in the lmc ; @xcite for the smc and m33 ; and for the solar neighbourhood and for the central region of m51 ) . from the theoretical side ,
the evolution and disruption of star clusters has been the subject of many studies using a variety of techniques to solve the gravitational @xmath0-body problem ( e.g. @xcite ) .
compared observationally determined disruption times in four galaxies to the corresponding times following from the @xmath0-body simulations of clusters in tidal fields of @xcite and .
they found that , based on the results of the simulations , the disruption time ( @xmath6 ) should depend on the cluster mass @xmath8 as @xmath25^{0.75 } \propto { \mbox{$m_{\rm c}$}}^{0.62}$ ] .
the index ( 0.62 ) , which is referred to as @xmath18 , was also determined from the observed age and mass distributions of different cluster populations @xcite and the mean value from the observations is @xmath26 .
the value of @xmath6 scales with the disruption time of a @xmath23 cluster as @xmath27 .
good agreement for @xmath28 between simulations and observations was found only for the star clusters in the smc . in the solar neighbourhood ,
the observed @xmath28 for open clusters ( @xmath29gyr , see ) is a factor of 5 lower than expected from the disruption time due to the galactic tidal field , combined with a realistic stellar mass function and with stellar evolution ( @xmath30gyr , see @xcite ) .
an even larger disagreement of almost an order of magnitude was found for @xmath6 of clusters in the central region of m51 .
the fact that the observed disruption times are shorter could be a result of time - dependent external perturbations , such as spiral arm passages and encounters with giant molecular clouds ( gmcs ) , which were ignored in the simulations of @xcite .
@xcite and @xcite showed that a single encounter with a massive gmc could disrupt a star cluster of a few hundred @xmath31 and @xcite showed that in the solar neighbourhood passages of spiral arms contribute significantly to the disruption of open clusters . in this study
we investigate the disruption of clusters by encounters with gmcs .
we want to quantify whether gmcs can explain the difference between the observed values of @xmath28 and the ones following from the simulations without the time - dependent external effects .
moreover , would simulations including encounters with gmcs preserve the numerical value of the index @xmath22 , where the observations and previous simulations already agree ?
the paper is organised as follows : in [ sec : initial ] we introduce the initial conditions for the clusters and gmcs and define the parameters involved in a cluster - gmc encounter . in
[ sec : analytical ] we give analytical calculations of the energy gain for a cluster due to various types of encounters .
these analytical formulae for the energy gain are verified with @xmath0-body simulations and compared to the mass loss of the cluster in [ sec : nbody ] .
the results of the @xmath0-body simulations are used in [ sec : disruption ] to derive an expression for the cluster disruption time in environments with different gmc densities .
the conclusions and discussion are presented in [ sec : conclusion ] .
in this study we consider encounters between molecular clouds and star clusters .
since 90% of the total molecular gas mass in our galaxy is in giant molecular clouds ( gmcs ) with mass @xmath32 @xcite and encounters with low mass clouds do not affect a star cluster much ( e.g. @xcite ) , we consider only the massive gmcs . in
[ sec : analytical ] we show that the cluster energy gain due to an encounter with a gmc scales with the gmc mass squared ( @xmath33 ) , supporting our assumption to consider only the massive clouds ( i.e. the gmcs ) .
@xcite showed that there is a relation between the size and mass of galactic gmcs of @xmath34 where @xmath35 is the mass of the gmc and @xmath36 is its radius '' ( from nebula or nuage ) to indicate the parameters of the gmc and `` @xmath37 '' for the cluster . ] .
the internal density profile is of the form @xmath38 .
[ eq : mrclouds ] implies a constant mean surface density of @xmath39 .
we use the spherically symmetric plummer model @xcite to describe the gmc , since it is mathematically convenient to use in analytical calculations ( [ sec : analytical ] ) .
the potential of this model is @xmath40 where @xmath41 is the gravitational constant , @xmath42 is the total mass and @xmath43 is the plummer radius . if we choose the plummer radius of a gmc equal to half the radius from eq .
[ eq : mrclouds ] , e.g. @xmath44 , the resulting profile is very similar to what would follow from a density profile of the form @xmath38 . in fig .
[ fig : pad ] we compare the descriptions for a plummer model ( eq . [ eq : plummer ] ) and constant surface density profile ( eq . [ eq : mrclouds ] ) for the potential in the top panel , the acceleration in the middle panel and the density profile in the bottom panel .
the plummer model has the advantage that the first derivative of the force is differentiable for each @xmath45 , which is necessary for the @xmath0-body simulations .
, following from eq .
[ eq : mrclouds ] and @xmath46 ( dashed line ) and for a plummer model with radius @xmath47 ( full line).,width=302 ] for the star cluster we also choose a plummer model ( eq . [ eq : plummer ] ) .
the number of particles is @xmath48 with a total mass of @xmath49 and all stars the same mass , resulting in a realistic mean stellar mass of 0.49(e.g .
the virial radius is set to 2 pc . for a plummer model
, the virial radius relates to the plummer radius of the cluster as @xmath50 .
the cluster half - mass radius relates to @xmath51 as @xmath52 .
therefore , the values of @xmath51 and @xmath9 in physical units are 1.18 pc and 1.54 pc , respectively .
the energy gain of a cluster due to an encounter with a gmc will be compared to the total initial cluster energy , e.g. the sum of kinetic and potential energy .
this is defined as @xmath53 , with @xmath54 , depending on the cluster model . for a plummer model @xmath55 , hence @xmath56 .
two particles with a relative velocity at infinity that is non - zero will experience a hyperbolic encounter .
the motion of the reduced mass particle is schematically represented in fig .
[ fig : impact_encounter ] .
the impact parameter and the velocity at infinity are referred to as @xmath57 and @xmath58 , respectively .
the distance of closest approach and the maximum velocity are called the encounter parameter @xmath59 and @xmath60 , respectively .
conservation of angular momentum relates @xmath58 and @xmath60 as @xmath61 consider a gmc with a plummer model according to eq .
[ eq : plummer ] and mass @xmath35 and a cluster with mass @xmath8 which is a point source . from conservation of energy and eq .
[ eq : v ] , it follows that @xmath57 and @xmath59 are related as @xmath62 note that @xmath57 is now a mixed function of encounter and impact variables . in the next sections we will show that the encounter parameter @xmath59 and @xmath60 determine the internal energy gain .
since we want to perform @xmath0-body simulations that result in an encounter with a certain desired @xmath59 and @xmath60 , we use eq .
[ eq : v ] and eq .
[ eq : pb ] to express @xmath59 and @xmath60 in terms of @xmath57 and @xmath58 .
in addition , in our simulations we will not start with the gmc at infinity , hence we will need to take into account the potential energy contribution at the beginning of the simulation . since the final expressions for @xmath57 and @xmath58 as a function of @xmath63 and @xmath8 are quite complex , but follow from simple mathematical exercises , we do not give them here .
there exists an analytical expression for the energy gain of a cluster due to an head - on encounter and due to a distant encounters . in the next section these relations
will be extended to the full range of encounter parameters , i.e. including close encounters .
the exact boundary between close and distant encounters is hard to define . here
we will distinguish three types of encounters : ( 1 ) head - on encounters for which @xmath64 ; ( 2 ) close encounters , where @xmath65 and ( 3 ) distant encounters , which is for @xmath66 . and
the velocity at infinity @xmath58 are denoted on the left side of the picture and the encounter parameter @xmath59 and maximum relative velocity @xmath60 are indicated at the point of closest approach.,width=302 ]
spitzer ( @xcite , hereafter s58 ) derived an expression for the energy gain of a cluster due to a distant encounter with a gmc .
he assumed that the gmc is a point mass and that the encounter velocity is much faster than the internal velocity of the stars in the cluster , which is referred to as the _ impulsive approximation_. the cluster energy gain can then be expressed in the encounter , gmc and cluster parameters as @xmath67 here @xmath68 is the mean - square position of the stars in the cluster and @xmath69 . in [ sec : nbody ] we show that @xmath59 and @xmath60 are the most convenient ones to use for calculating the energy gain . ] .
@xcite used combined hydro - dynamical and @xmath0-body calculations of a gmc - cluster encounter , to show that eq .
[ eq : despitzer ] is very accurate .
he also notes the inconvenience of the ill defined parameter . in principle @xmath68
follows from the cluster profile , but this variable is very sensitive to variations of the positions of stars far from the cluster centre , hence it is not constant during the simulation because of scattering of stars out of the cluster . even in simulations covering a short time span , @xmath68 will increase due to stars that get at larger distances from the centre due to 2-body encounters in the core .
@xcite have shown with @xmath0-body calculations of encounters between two plummer models with radii @xmath70 and @xmath71 , that the predictions of eq .
[ eq : despitzer ] hold when @xmath72 . in [ sec : nbody ] we examine with @xmath0-body simulations in what regime the results of eq .
[ eq : despitzer ] are valid , and at what moment during the encounter the value for should be taken .
gravitational focusing will deflect many cluster orbits which start with @xmath73 to encounters with @xmath74 , for which the relation of s58 ( eq . [ eq : despitzer ] ) does not hold . to obtain an expression for the energy gain for close encounters
, we first explore the opposite of distant encounters , namely head - on encounters ( i.e. @xmath64 ) . in those encounters
the extent of the gmc has to be taken into account , since assuming a point mass gmc would severely overestimate the energy gain .
the energy gain in close encounters should be an interpolation between the distant encounter and head - on results . following @xcite ( chap .
7.2 , hereafter bt87 ) we can derive the cluster energy gain for an head - on encounter between a cluster and a gmc , both described by a plummer model .
let @xmath75 be cylindrical coordinates , with the @xmath76-axis coinciding with the gmc trajectory .
due to symmetry , the stars in the cluster get a velocity increase only perpendicular to the trajectory of the gmc , of size @xmath77 where analogous to the case of the distant encounters we use @xmath60 for the relative velocity between the cluster and the gmc .
[ eq : dv ] can be used to calculate the energy gain of the cluster : @xmath78 ^ 2\,\sigma(r)\,r\,{\mbox{${\rm d}$}}r.\ ] ] here @xmath79 is the surface density of the cluster obtained by projecting parallel to the line of motion of the gmc . since the cluster is described by a plummer model , @xmath79 is @xmath80 so the energy gain of the cluster due to an encounter between two plummer models is @xmath81 the result of the integration is a function of @xmath51 and @xmath82 and is equal to @xmath83 when @xmath84 . in that case , eq .
[ eq : de1 ] reduces to the result found by bt87 : @xmath85 when @xmath86 , the result of the integration is a function of @xmath51 and @xmath82 . introducing a new variable @xmath87 , the ratio of the gmc and cluster radii , eq .
[ eq : de1 ] can be written as @xmath88 where in the first factor we recognise the result for @xmath89 ( e.g. eq .
[ eq : debt ] ) and the rest is a correction factor @xmath90 which is only a function of @xmath91 . we can write a more general expression for the energy gain of head - on encounters @xmath92 with @xmath93 the expression for @xmath90 is rather complicated , but for realistic values of @xmath91 ( i.e. @xmath94 ) , @xmath90 can be approximated with high accuracy by @xmath95 ( see fig .
[ fig : head_on ] ) . with eq .
[ eq : deheadon ] and eq .
[ eq : cx ] we have a general expression for the energy gain of head - on gmc - cluster encounters , when both are described by plummer models , valid for any radius and mass of the cluster and the gmc . in
[ subsec : headon_nbody ] we will confront eq .
[ eq : deheadon ] and eq .
[ eq : cx ] with results from @xmath0-body simulations . the energy gain for a close encounter ,
i.e. an encounter for which the cluster moves through the cloud within a few @xmath82 from the gmc centre , should be a smooth interpolation between the result for a head - on ( eq .
[ eq : deheadon ] ) and distant encounters ( eq . [ eq : despitzer ] ) .
when replacing @xmath96 in eq .
[ eq : despitzer ] by @xmath97 , we get the result for head - on encounter ( eq .
[ eq : deheadon ] ) .
one way of connecting the two is to add a term in the denominator of eq .
[ eq : despitzer ] , preventing the result to diverge to infinity at @xmath64 , so that it converges to eq .
[ eq : deheadon ] , is @xmath98 if we use the approximation given in
[ subsec : headon ] for @xmath90 ( i.e. @xmath99 ) and the fact that @xmath100 , @xmath101 and the real gmc radius @xmath36 relates to @xmath82 as @xmath102 , we can rewrite eq .
[ eq : detot ] as @xmath103 in the next section , this relation will be compared to results of @xmath0-body simulations for encounters with @xmath59 ranging from 0 to @xmath104 .
the @xmath0-body calculations were carried out by the @xmath105 integrator , which is part of the starlab software environment ( @xcite ; @xcite ) .
@xmath106 uses a fourth - order hermite scheme and includes special treatments of close two - body and multiple encounters of arbitrary complexity . the special purpose grape-6 systems @xcite of the university of amsterdam are used to accelerate the calculation of gravitational forces between stars .
the potential , acceleration and jerk of the stars due to the gmcs are derived from eq .
[ eq : plummer ] and calculated for each star individually , based on its position and velocity with respect to the gmc .
we assume the gmcs are moving potentials which do not interact .
indeed , we are interested in the star cluster and since we consider gmcs that are @xmath107 times more massive than the cluster , the gravitational pull from the star cluster on the gmcs is negligible and will therefore not affect its orbit .
although gmcs have finite life times ( @xmath108myr ) , the encounter durations are sufficiently small ( @xmath109myr ) to not be affected by this .
[ cols= " > , > , > , > , > , > , > , > " , ] [ tab : runs ] we want to avoid sudden jumps in the contribution of the gmc to the cluster potential energy at the beginning and at the end of the simulations . in principle , our gmc should start at infinity , have an encounter with the star cluster and then continue until infinity . since we do not want to evolve the cluster to long before the encounter , to restrict simulation time and to avoid too much internal dynamical evolution , we put the gmc initially at a distance of @xmath110 from the cluster and give it zero mass . while the gmc travels from @xmath111 to @xmath112 , approaching the cluster , we increase its mass gradually from 0 to @xmath35 . similarly , after the encounter and while the gmc travels from @xmath112 to @xmath111 we decrease its mass gradually from @xmath35 to 0 . to give the cluster time to get back to equilibrium and to give stars with velocities higher than the escape velocity time to leave the cluster we evolve it further for a time equal to half that elapsed after the encounter . to test the different analytical expressions of
[ sec : analytical ] for the energy gain of the cluster , we perform series of head - on ( @xmath64 ) , distant ( @xmath113 ) and close encounters with different gmc radii and encounter parameters . the initial velocity ( @xmath58 ) and impact parameter ( @xmath57 ) of the gmc are calculated from the desired values of @xmath59 and @xmath60 , using the expressions for gravitational focusing of
[ subsec : relations ] .
the cluster properties are described in
[ subsec : cluster ] .
a summary of all runs is given in table [ tab : runs ] .
the results will be discussed in
[ subsec : headon_nbody]-[subsec : dedm ] .
all clusters are scaled to @xmath0-body units , such that @xmath114 and @xmath115 , following @xcite .
then @xmath116 is the unit of length and follows from the scaling of the energy , since @xmath117 , where @xmath118 is the total ( potential and kinetic ) initial energy of the cluster .
we assume virial equilibrium at the start of the simulation , which implies @xmath119 .
we perform @xmath0-body simulations of head - on encounters between a cluster and a gmc , both described by a plummer model .
gmcs with different plummer radii @xmath82 were used to compare simulations to the analytical result of eq .
[ eq : deheadon ] .
the energy gain of the cluster is always expressed in units of the initial cluster energy , i.e. the fractional energy gain ( @xmath120 ) .
since we are here interested in the effect of the gmc radius on , we fix the mass of the gmc at a constant value .
the results can then be compared with the predicted value from eq .
[ eq : deheadon ] .
more specific , when we divide the energy gain for the different cases by the value that was predicted by bt98 ( e.g. eq .
[ eq : debt ] ) , we should get the factor @xmath90 of eq .
[ eq : cx ] . in [ subsec : dedm ]
we show that there is no good agreement between the simulations and the analytical predictions when @xmath121 . to keep @xmath122 for all gmc radii , we choose @xmath123 , which is somewhat lower than a realistic value for .
we choose integer values for @xmath91 from 1 to 10 , where eq .
[ eq : debt ] and eq .
[ eq : deheadon ] should be valid for @xmath124 and @xmath125 , respectively .
the encounter velocity for all simulations is @xmath126 , where @xmath127 is the internal velocity dispersion of stars in the cluster . in fig .
[ fig : head_on ] we show the results of from the @xmath0-body simulations for different @xmath91 with diamond symbols , the value of bt87 with a filled circle ( left ) and the result of eq .
[ eq : deheadon ] with the full line .
the latter relation is also shown with a dashed line , but with an approximation for @xmath90 of the form @xmath128 .
this approximation is very close to the real @xmath90 for @xmath129 .
excellent agreement between the analytical work and the @xmath0-body results is found in all cases . in fig .
[ fig : snap_head_on ] we show snapshots of a cluster during a head - on encounter with a gmc . in the top panels the line of sight is perpendicular to the trajectory of the gmc , whose direction is parallel to the @xmath130-axis .
the attractive force of the gmc parallel to its trajectory before and after the encounter cancel out , resulting in displacement of stars only perpendicular to the line of motion of the gmc . in the bottom panels the line of sight is along the motion of the gmc ,
and it can be seen that the stars escape and form a disk like structure of unbound particles around the cluster .
the time with respect to the encounter moment is indicated in each panel in the top row , in units of @xmath131 , where @xmath132 . for different @xmath87 and with @xmath133 .
the analytical approximation of bt87 ( eq . [ eq : debt ] ) for head - on encounters between identical plummer models is indicated with a filled circle ( left ) .
the @xmath0-body results are shown with diamonds . the analytical prediction of eq .
[ eq : deheadon ] with the correction factor @xmath90 for different cloud radii is shown with the full line .
the predicted energy gain with an approximation for @xmath90 ( i.e. @xmath128 ) is shown with the dashed line.,width=302 ] for the simulations of distant encounters we choose @xmath134 .
the tidal forces of the gmc deform the cluster into a cigar like shape , as can be seen in fig .
[ fig : snap_distant ] . in fig .
[ fig : de_example ] we show with the full lines the evolution of the internal energy of the cluster for encounters with three different values for @xmath59 as a function of time , in units of the encounter time ( @xmath135 ) .
( note that for head - on encounters @xmath132 . )
we define time ( @xmath136 ) such that the moment of closest approach is at @xmath137 .
the value of for these runs is @xmath138 and the gmc radius is equal to the cluster radius .
the gmc radius is chosen small in this example in order not to have too large values of @xmath59 when the ratio @xmath139 is large .
otherwise , the energy gain would be too small .
the results of eq .
[ eq : despitzer ] for these encounters are overplotted as the dashed lines .
the value of changes during the simulations due to internal dynamical effects and because of deformation of the cluster by the gmc .
this last effect is clearly visible in figs .
[ fig : snap_head_on ] and [ fig : snap_distant ] . since the result of eq .
[ eq : despitzer ] is very sensitive to the value of , we performed tests to find at which time should be taken to get the best agreement between eq .
[ eq : despitzer ] and the simulations .
if we take this value too early , eq . [ eq : despitzer ] will underestimate the energy gain compared to the result of the simulation .
this because dynamical evolution will increase . when using a value which is determined during the encounter ( i.e. @xmath140 ) , the energy gain from eq .
[ eq : despitzer ] overestimates the true value .
we found that best agreement was found when using at @xmath141 . from fig .
[ fig : de_example ] it can be seen that this is moment where the energy of the cluster starts to increase , whereas the cluster energy is almost constant before that moment .
the variations of the relative velocity between the cluster and the gmc are smaller than the variations in , and for that reason the result of eq .
[ eq : despitzer ] is less sensitive to the time that we choose the velocity . since most of the energy is injected between @xmath142 and @xmath143 ( fig .
[ fig : de_example ] ) , the best agreement is obtained when the relative velocity at @xmath137 , e.g. @xmath60 , is used . ,
@xmath144 and three different impact parameter @xmath59 .
the dashed lines indicate the final energy gains predicted by s58 ( eq . [ eq : despitzer]).,width=302 ] to test the validity of the interpolation between the distant and the head - on encounters ( eq . [ eq : detot ] ) , we performed @xmath0-body simulations of various encounters with impact parameter @xmath59 ranging from 0 to 10 times the gmc radius .
the encounter velocity was chosen @xmath145 , such that the fractional energy gain is always lower than 1 ( eq .
[ eq : deheadon ] ) . the impact parameter @xmath57 and initial velocity @xmath58 of the gmc where calculated using eq .
[ eq : pb ] .
the gmc mass is @xmath146 and we used the observed mass - radius relation for gmcs ( eq.[eq : mrclouds ] ) , resulting in @xmath147 and therefore @xmath148 ( eq . [ eq : cx ] ) . in fig .
[ fig : de_p ] we show the result of from the @xmath0-body simulations ( diamonds ) and the analytical prediction of eq .
[ eq : detot ] ( full line ) .
the predictions for the two extreme values of @xmath59 , i.e. for distant and for head - on encounters are shown with the dashed line and the filled circle ( left ) , respectively .
the analytical expression eq .
[ eq : detot ] agrees very well with the results from the @xmath0-body simulations .
it can be seen that the prediction of s58 is accurate for @xmath149 , in agreement with what was found by @xcite .
note that in fig .
[ fig : de_p ] we have used as a scale length to normalise @xmath59 .
this is possible since gmcs have finite sizes .
if we were dealing with objects of infinitesiaml size , as e.g. black holes , this scaling would not be possible . thus , fig .
[ fig : de_p ] can not be applied to those cases . for large @xmath59 ( @xmath150 )
the energy gain only depends on @xmath59 .
this is demonstrated by the good agreement with the point source approximation of s58 , since for these cases @xmath151 . and for @xmath152 .
the results of the @xmath0-body simulations are shown with diamonds .
the analytical result for head - on encounters , including the correction for the radius of the gmc is shown with a filled circle ( left ) .
the result for distant encounters from s58 is shown with the dashed line .
the result from the interpolation between the two analytical predictions ( e.g. eq . [ eq : despitzer ] and [ eq : deheadon ] ) , i.e. eq . [ eq : detot ] , is shown with the full line.,width=302 ] in units of and for different values of @xmath60 .
the results of the @xmath0-body simulations are shown with diamonds , where the size of the diamonds corresponds to the encounter velocity .
the analytical results from eq .
[ eq : detot ] are shown with the full lines.,width=302 ] to see if the agreement between our predicted energy gain from eq .
[ eq : detot ] and @xmath0-body simulations holds for lower velocities , we performed a similar series @xmath0-body simulations as shown in fig .
[ fig : de_p ] for @xmath153 and @xmath154 .
the values of for these three different values of and different values of @xmath59 is shown in fig .
[ fig : de_v ] . for the lower velocities
we again find very good agreement .
however , when the fractional energy gain is larger than 1 , the @xmath0-body simulations show a smaller energy gain than the predictions . the difference is larger for larger energy gains .
similar results were obtained by @xcite for one - dimensional shocks due to spiral arms .
this is probably because the impulsive approximation does not hold anymore .
when the gmc is very massive , or when the relative velocity is low , the stars in the cluster are displaced too much before the encounter .
the symmetry assumption of the encounter of the impulse approximation then does not hold anymore .
the energy gain from the encounter is absorbed mainly by the stars far from the cluster centre , which can gain velocities much higher than the escape velocity .
a bound core will remain , even when @xmath121 .
both topics will be discussed in more detail in [ subsec : dedm ] . in the previous section we found good agreement between the @xmath0-body simulations and our analytical formula for of a cluster due to encounters with gmcs with different @xmath59 and @xmath60 as long as @xmath122 ( see eq .
[ eq : detot ] ) . in fig .
[ fig : de_v ] we show that deviations occur whenever the energy gain is larger than the initial cluster energy .
we did a few more simulations of head - on encounters with higher gmc masses and compare all the predicted energy gains to the results from the simulations .
the results are shown in fig .
[ fig : dde ] . a one - to - one relation ( dashed line ) holds until @xmath155 .
a simple functional ( @xmath156 $ ] ) , which starts to deviate from a one to one relation around @xmath157 and levels off at @xmath158 is also shown in fig .
[ fig : dde ] and describes the results well .
( circles ) . a one - to - one relation is overplotted as the dashed line . a simple functional form , describing a linear function that saturates at a value of 12 , fits the result at the high energy part well.,width=321 ] to derive a realistic disruption time of clusters due to encounters with gmcs , we need to know how the energy gain relates to the number of stars , or mass , that is lost from the cluster in the different encounters . in fig .
[ fig : dedm ] we show the fractional mass loss ( @xmath159 ) as a function of @xmath160 .
the dashed line indicates a one - to - one relation and the full line is a fit to the results .
the grey areas indicate the regions where @xmath159 can not be defined , since the cluster can not loose less then one star ( i.e. @xmath161 , lower region ) , and it can not loose more than the total initial cluster mass ( i.e. @xmath162 , upper region ) .
the fit shows an almost linear relation , where is a fraction @xmath163 of .
given the good agreement between the predicted energy gain and the one from the simulations , we can conclude that a fraction @xmath163 of the input energy is used to unbind stars .
this implies that a fraction @xmath164 of went into the velocity of the escaping stars , so that these have escaped with velocities much higher than the local escape velocity .
combined with the saturation of fig .
[ fig : dde ] , this shows that some encounters with @xmath165 , called overkill encounters @xcite , can completely disrupt a cluster .
we will further discuss this topic in [ subsec : tdissingle ] . since we consider isolated clusters
, one might argue that there can be stars that have remained bound , but that would have been pushed over the tidal radius if the cluster was located in a galaxy . assuming a disk galaxy with a flat rotation curve , with rotational velocity of @xmath166 , the tidal radius of a cluster depends then on its mass and galactocentric distance ( @xmath167 ) as @xmath168 .
the estimated values of @xmath169 for our cluster at @xmath170\times r_0 $ ] , where @xmath171kpc , i.e. the solar neighbourhood , are @xmath172\,$]pc , respectively .
we counted for our simulations the number of extra stars ( @xmath173 ) which after the encounter beyond these three radii .
there is a linear relation between and , where the ratio of the two quantities is @xmath174 $ ] for the three different tidal limits , respectively .
so the effective value of @xmath175 is somewhat higher when we take into account the tidal limit of the cluster . for the solar neighbourhood
a good estimate would be @xmath176 . for clusters in the centres of galaxies
the tidal radius is small and @xmath175 can be a factor @xmath177 higher . from now on we will assume a linear relation @xmath178 , with @xmath179 , which takes into account stars that have become unbound in isolation and/or are pushed over the tidal limit .
this will be used in [ sec : disruption ] to derive the mass loss of clusters in environment with gmcs and the resulting cluster disruption time .
hitherto , the cluster disruption time has always been defined as the time needed to bring the cluster energy to zero by periodically injecting energy with shocks ( e.g. @xcite for the disruption time by disk shocks and @xcite for the disruption by gmcs ) .
we have shown that this overestimates the disruption time by a factor of @xmath180 .
a more accurate disruption time , taking into account this factor @xmath175 is derived in the following section . ) and the fractional energy gain ( @xmath160 ) for the simulations of [ sec : nbody ] ( circles ) .
the dashed line shows a one - to - one relation and the full line is a fit to the simulations .
the grey areas indicate the regime where @xmath181 can not be defined .
, width=321 ]
with the description of the mass loss from a cluster due to different types of encounters with gmcs of
[ sec : analytical],[sec : nbody ] we can derive the disruption time of star clusters in an environment with gmcs , such as the solar neighbourhood . when considering a distribution of gmcs and clusters with a relative velocity dispersion @xmath182 , the probability of a relative velocity in the range @xmath183 can be expressed as ( bt87 , eq .
7 - 60 ) @xmath184 assuming a spatially homogeneous distribution of gmcs , the rate of encounters in the range @xmath185 is then @xmath186 where @xmath187 is the number density of clouds .
the energy of the cluster increases at a rate @xmath188 where @xmath189 follows from eq .
[ eq : detot ] after converting @xmath57 and @xmath58 to @xmath59 and @xmath60 due to gravitational focusing .
this integral can be calculated numerically . in fig .
[ fig : encounter_rate ] we illustrate this using the parameters for clusters and gmcs used in the @xmath0-body simulations of
[ sec : nbody ] , e.g. @xmath134 , = 6.8pc and @xmath190pc .
we assume that young clusters and gmcs have the same velocity dispersion , i.e. @xmath191 . the velocity dispersion of the motion of gmcs with respect to clusters , is then @xmath192 . in the left panel of fig .
[ fig : encounter_rate ] we show the impact rate @xmath193 of eq .
[ eq : encounterrate ] in units of @xmath187 as a function of @xmath57 and @xmath58 , in units of @xmath82 and @xmath194 , respectively . in the middle panel
we show for individual encounters as a function of @xmath195 and @xmath196 , where we have converted @xmath57 and @xmath58 to @xmath59 and @xmath58 first , and then substituted these values in the equation for energy gain ( eq . [ eq : detot ] ) .
the rate of energy gain as a function of @xmath195 and @xmath196 is the product of the encounter rate times the energy gain per encounter and is shown in the right panel .
the isocontours indicate the regions where 10 , 50 and 75% of the total energy per unit time is produced . this total energy was not determined from this graph , but from a numeric integration of a similar figure , but with an area that is 100 times bigger .
most of the energy is injected by fast ( @xmath197 ) and close ( @xmath198 ) encounters .
the encounter velocities ( @xmath60 ) are even higher than @xmath199 due to gravitational focusing and this means that the impulse approximation is valid for these encounters . in
[ subsec : allp ] we found that the distant encounter regime of s58 holds for @xmath149 .
the peak in the right panel of fig .
[ fig : encounter_rate ] around @xmath200 illustrates that the approximations of s58 for a point source gmc are not applicable for the encounters that inject the majority of the energy . in that regime
it is important to use the more accurate description for the energy gain found in
[ subsec : detotal ] ( eq .
[ eq : detot ] ) .
bt87 solved eq .
[ eq : energyrate ] by substituting the energy gain for distant encounters of s58 .
they then integrate @xmath57 from @xmath36 to infinity and correct for the missing close encounters with a factor @xmath201 of order unity .
the advantage of this solution is that the result has @xmath202 in the denominator , which can be combined with @xmath35 and @xmath187 to express the final result in the observable molecular gas density ( @xmath14 ) and surface density of individual gmcs ( @xmath13 ) .
our results allow us to quantify the parameter @xmath201 of bt87 by integrating eq .
[ eq : energyrate ] numerically from 0 to infinity and from @xmath36 to infinity .
when gravitational focusing is ignored , the value of @xmath201 depends only on @xmath35 , since in that case @xmath203 and @xmath204 . including gravitational focusing makes @xmath201 dependent on @xmath194 as well ,
since then encounters with initial @xmath205 can result in @xmath206 .
intuitively one thinks that the inclusion of gravitational focusing will increase the value of @xmath201 , since there are more encounters with @xmath206 to correct for .
however , the encounter velocity is always higher than @xmath58 , which will make the energy gain of the cluster for these encounters lower . in fig .
[ fig : g ] we show the resulting @xmath201 for various values of @xmath35 ignoring gravitational focusing ( full line ) and for three values of @xmath194 .
the horizontal dashed line indicates the estimated value of bt87 for @xmath201 . for the solar neighbourhood (
e.g. @xmath207 ) our results show that @xmath208 , slightly lower than predicted by bt87 . for close encounters ( see text for details ) as a function of gmc mass .
the full line shows the result when gravitational focusing is ignored , and the other lines show the results for 3 different values of @xmath194 when gravitational focusing is considered .
the predicted value of @xmath209 of bt87 is indicated with the dashed horizontal line.,width=321 ] using the relation between energy gain and mass loss of
[ subsec : dedm ] we can define the mass loss rate of the cluster as @xmath210 where @xmath211 is a dimensionless constant that relates fractional energy gain to fractional mass loss ( see fig . [
fig : dedm ] and [ subsec : dedm ] ) .
the initial cluster energy is @xmath212 , where @xmath54 , depending on the cluster profile @xcite .
we can derive a relation for the mass loss , using the expression for @xmath213 of bt87 ( their eq .
7 - 66 ) , where we substitute @xmath214 , @xmath215 and @xmath216 @xmath217 where the ratio @xmath218 depends on the cluster model . for the plummer cluster that we used @xmath219 and
for king profiles with dimensionless potential depth values of @xmath220 $ ] it follows that @xmath221 $ ] . from eq .
[ eq : mdot ] we can derive the disruption time of clusters , by defining @xmath222 for the constants in eq .
[ eq : tdis ] we have to make several assumption .
first , we assume @xmath211 , as was found in [ subsec : dedm ] .
for @xmath194 we take @xmath223 , based on observations of gmcs in the solar neighbourhood @xcite and open clusters .
for the cluster profile we assume a king profile with @xmath224 , corresponding to @xmath225 , which is the mean concentration of the galactic open clusters @xcite .
king profiles are more realistic due to the tidal truncation , which the plummer model does not have . for the @xmath226 profile considered here , @xmath227 and @xmath228 ( bt87 ) . the solution for @xmath6 of eq .
[ eq : tdis ] is very sensitive to the choice of @xmath9 .
observations of ( young ) extra - galactic star clusters show that the projected half - light radius ( @xmath229 ) of the majority of star clusters is confined to a narrow range of 2 - 4 pc ( e.g. ) and a very shallow relation between the radius and the mass .
a factor of two difference in radius and nearly equal masses still results in a factor of eight difference in .
this implies that clusters that are a few times larger than the mean radius are very unlikely to survive and cluster that are a few times smaller will be almost insensitive to encounters by gmcs .
we choose the use the recently observed relation between radius and mass of , who fits a function of the form @xmath230 to a sample of young star clusters in nearby spiral galaxies and finds that @xmath231 and @xmath232 .
there is a large scatter around the fit relation , which , combined with the high sensitivity of on ( eq .
[ eq : tdis ] ) , makes it hard to relate mass to radius .
however , since the relation of reflects the median relation between and , the conversion will probably hold for a large cluster population . to convert @xmath229 to @xmath9 we use the relation @xmath233 @xcite and with
that we can rewrite the result of as @xmath234 with @xmath20 . substituting eq .
[ eq : rm ] in eq .
[ eq : tdis ] and inserting all constants , yields an expression for the disruption time that is independent of @xmath9 and depends on the assumed index that relates @xmath9 and @xmath8 @xmath235 with @xmath19 .
we normalised the product @xmath236 to the value for the solar neighbourhood ( @xmath237pc@xmath238 and @xmath239pc@xmath240 , @xcite ) .
when taking the 1 sigma upper bound value of @xmath241 from , the index @xmath242 agrees very well with the value of 0.62 found from observations and from @xmath0-body simulations of clusters in the solar neighbourhood , dissolving under the combined effect of a tidal field , a realistic mass function and stellar evolution @xcite .
these simulations adopted a slightly different relation between and since they assumed that at a given galactocentric distance the cluster size scales with the tidal radius ( @xmath243 ) .
the multiplicative constant of eq .
[ eq : tdis2 ] is a factor of about 7 higher than what was found from earlier studies ( e.g. s58 ; @xcite ; bt87 ) .
the difference can largely be explained by our factor @xmath4 from [ subsec : dedm ] , that relates energy gain to mass loss .
the remaining difference is caused by the lower value we found for @xmath201 and a slightly different value for the product @xmath236 . in the derivation of in the previous sections we integrated over the full range of @xmath57 and @xmath58 in eq .
[ eq : energyrate ] to get an expression for the rate of energy gain of a cluster due to gmc encounters .
doing so , we implicitely assume that disruption is caused by a large number of encounters . however , if the cluster is completely disrupted by the first encounter with a gmc , eq .
[ eq : tdis ] will overestimate .
[ fig : dedm ] shows that a single encounter can indeed disrupt the cluster ( see also @xcite and @xcite ) .
if cluster disruption was always caused by single encounters , then eqs . [ eq : tdis]&[eq : tdis2 ] would overestimate the disruption time in all cases .
the most accurate approach is to include the saturation function of fig .
[ fig : dde ] in the relation between energy gain and mass loss ( eq .
[ eq : rel_dm_de ] ) , i.e. to exclude encounters with @xmath244 .
this results in = 1 for @xmath245 .
however , this would not have allowed us to come to the convenient linear dependence of on and in eq .
[ eq : tdis ] .
alternatively , we here estimate the typical time scale for a cluster to be disrupted by one single encounter ( @xmath246 ) .
following @xcite , we can calculate the critical encounter distance @xmath247 for which one encounter will result in total disruption of the cluster .
based on eq .
[ eq : detotrh ] , on the expression for the cluster energy @xmath248 and on the relation between and from [ subsec : dedm ] we can derive by solving for @xmath59 in @xmath249 , such that @xmath250 in here @xmath251 , which is the term in the denominator of eq . [ eq : detotrh ] that makes @xmath252
converge to the correct value for head - on encounters ( @xmath64 ) .
@xcite did not include this term , since he assumed point - mass gmcs ( eq . [ eq : despitzer ] ) .
( note that eq .
[ eq : detotrh ] is equal to eq .
[ eq : despitzer ] when @xmath253 ) .
the number of encounters with encounter parameter @xmath59 depends on the gmc number density ( @xmath187 ) as @xmath254 we can solve for @xmath255 by setting @xmath256 and @xmath257 , which is then the time it takes for a disruptive encounter to occur @xmath258 in here gravitational focusing is ignored . following @xcite and assuming point - mass gmcs , e.g. @xmath253 , we find an expression of of the form @xmath259 where we have used @xmath260 again . at first sight
this relation for the disruption time looks similar to what we derived in the previous section ( eq .
[ eq : tdis ] ) .
the main difference is the scaling with the cluster density ( @xmath261 ) . in fig .
[ fig : tdissingle ] we show a comparison between of eq . [ eq : tdissingle ] ( dotted line ) and of eq . [ eq : tdis ] ( full line ) for different .
here we used a constant = 3.75 pc and the same values for @xmath262 and @xmath263 as in [ sec : disruption ] .
the value of is smaller than the result of eq .
[ eq : tdis2 ] for clusters with @xmath264 .
this implies that for massive clusters the life time is limited by the first disruptive encounter , rather than the successive heating by multiple encounters . for cluster of lower mass ,
the time it takes to disrupt by encounters with @xmath265 is shorter than .
this mass dependence is counter intuitive at first , but is explained by the difference in scaling with @xmath8 ( eqs .
[ eq : tdis]&[eq : tdissingle ] ) .
note that the crossing point ( fig .
[ fig : tdissingle ] ) does not depend on , since eqs .
[ eq : tdis]&[eq : tdissingle ] depend in the same way on . in fig .
[ fig : tdissingle ] we also plot the result of by substituting eq .
[ eq : pcrit ] with @xmath266 in eq .
[ eq : tdissingle1 ] , based on @xmath134 ( dashed line ) and @xmath267 ( dotted - dashed line ) and eq . [ eq : mrclouds ] .
here we use the same value for , and then assume that all mass is in clouds with @xmath134 and @xmath268 , respectively , to solve for @xmath187 ( eq . [ eq : tdissingle1 ] ) . for low mass clusters , these solutions for agre with the result of @xcite , since there @xmath269 , where the assumption of point - mass gmcs is valid ( see fig [ fig : de_p ] ) . for clusters with @xmath270 , the solution for is higher than that of the point - mass assumption .
this is because is smaller when extended gmcs are considered . in close encounters
the extent of the gmc , e.g. @xmath271 , is more important ( eq .
[ eq : detot ] ) .
there is no solution for clusters with @xmath272 , meaning that it is not possible to disrupt these clusters by a single encounters with a gmc with masses up to few times @xmath268 .
the fact that the solutions for for realistic ( extended ) gmcs are always larger than from eq .
[ eq : tdis ] justifies our assumptions of
[ subsec : environment ] . however , fig .
[ fig : tdissingle ] shows that for a cluster with @xmath273 the two solutions are close , meaning that clusters are destroyed by just a few encounters .
this means that individual clusters will have strongly varying lifetimes .
our results of [ sec : disruption ] will , however , be valid as a statistical mean .
( full line ) and the single encounter disruption time ( @xmath246 ) .
the dotted line shows the result of @xcite for point - mass gmcs and the dashed and dotted - dashed lines show the results for extended gmcs with @xmath134 and @xmath268 , respectively.,width=302 ]
we have derive an expression for the cluster disruption time ( @xmath6 ) due to encounters with gmcs , based on the results of @xmath0-body simulations of different types of encounters between a cluster and a gmc .
the expression for ( eq . [ eq : tdis2 ] ) allows us to estimate the disruptive effect of gmcs in different environments and to compare it to other effects such as the galactic tidal field , stellar evolution and spiral arm shocks . from eq .
[ eq : tdis2 ] we see that the disruption time for a @xmath23 cluster in the solar neighbourhood is @xmath274gyr .
this is a factor of 3.5 shorter than what was found by @xcite for the combined effect of a realistic mass function , stellar evolution and the galactic tidal field ( @xmath275gyr ) .
note that the index @xmath18 is comparable , if we take the upper bound value of @xmath232 , since @xcite found @xmath276^{0.75}$ ] . showed that @xmath277^{0.75}\propto { \mbox{$m_{\rm c}$}}^{0.62}$ ] .
the same mass dependence was also found from the age and mass distribution of cluster samples in different galaxies by @xcite and from the age distribution of clusters in the solar neighbourhood by .
for this last sample they found a shorter scaling value of @xmath29gyr .
[ eq : tdis2 ] combined with the observed relation between @xmath9 and @xmath8 ( eq . [ eq : rm ] ) suggests that the combined mass loss by gmc encounters and evaporation in a tidal field preserves the index @xmath22 . a similar scaling between @xmath6 and @xmath8 was found for the disruption time by spiral arm passages .
the combination of all these effects and a comparison to the observed age distribution of open clusters in the solar neighbourhood will be discussed in more detail by @xcite .
the result of eq .
[ eq : tdis2 ] can easily be applied to environments / galaxies with different molecular gas densities ( @xmath278 ) .
the mid - plane density ( @xmath14 ) of gmcs scales with @xmath278 as @xmath279 , where @xmath280 is the vertical scale length of molecular gas , which is around @xmath281pc in the solar neighbourhood @xcite . assuming that @xmath280 is the same in m51 , and therefore @xmath282pc
, we can predict @xmath28 due to gmcs encounters once we know @xmath278 in m51 and compare this to the value of @xmath28 for clusters in m51 .
find @xmath283pc@xmath238 for the central region of m51 , which is 14 times higher than in the solar neighbourhood @xcite .
this implies @xmath284myr .
note that this value is an upper limit , since we have assumed that the surface density of individual gmcs ( @xmath13 ) is independent of @xmath278 , while this could be higher in environments with higher @xmath278 .
in addition , the values of @xmath175 and @xmath201 in eq .
[ eq : tdis ] will probably be higher in the centre of m51 ( see [ subsec : dedm ] and [ sec : disruption ] ) , which will make shorter .
also , we have assumed the same density profile as for the galactic open cluster , e.g. @xmath285 . a density profile of @xmath286 , as found for the galactic globular clusters , would reduce by a factor of two ( eq .
[ sec : disruption ] ) .
the resulting value of @xmath28 , however , is very close to the value of @xmath287myr , determined observationally by from the age and mass distribution of star clusters in m51 .
based on this work we think that we can explain the short disruption times of clusters in the centre of m51 , if gmc encounters are the dominant disruption effect .
we thank the referee , roland wielen , for useful comments and pointing us to the problem of single disruptive encounters ( [ subsec : tdissingle ] ) .
we enjoyed discussions with nate bastian , sren larsen and sverre aarseth .
this work was done with financial support of the royal netherlands academy of arts and sciences ( knaw ) , the dutch research school for astrophysics ( nova , grant 10.10.1.11 to hjglm lamers ) and the netherlands organisation for scientific research ( nwo , grant 635.000.001 ) .
the simulations were done on the modesta platform in amsterdam .
mg thanks the sterrekundig instituut `` anton pannenkoek '' of the university of amsterdam for their hospitality during many pleasant visits . | we investigate encounters between giant molecular clouds ( gmcs ) and star clusters . we propose a single expression for the energy gain of a cluster due to an encounter with a gmc , valid for all encounter distances and gmc properties .
this relation is verified with @xmath0-body simulations of cluster - gmc encounters , where the gmc is represented by a moving analytical potential .
excellent agreement is found between the simulations and the analytical work for fractional energy gains of @xmath1 , where @xmath2 is the initial total cluster energy . the fractional mass loss from the cluster scales with the fractional energy gain as @xmath3 , where @xmath4 .
this is because a fraction @xmath5 of the injected energy goes to the velocities of escaping stars , that are higher than the escape velocity .
we therefore suggest that the disruption time of clusters , @xmath6 , is best defined as the time needed to bring the cluster mass to zero , instead of the time needed to inject the initial cluster energy .
we derive an expression for @xmath6 based on the mass loss from the simulations , taking into account the effect of gravitational focusing by the gmc .
assuming spatially homogeneous distributions of clusters and gmcs with a relative velocity dispersion of , we find that clusters loose most of their mass in relatively close encounters with high relative velocities ( @xmath7 ) .
the disruption time depends on the cluster mass ( @xmath8 ) and half - mass radius ( @xmath9 ) as @xmath10 , with @xmath11 for the solar neighbourhood and @xmath12 scales with the surface density of individual gmcs ( @xmath13 ) and the global gmc density ( @xmath14 ) as @xmath15 .
combined with the observed relation between and , i.e. @xmath16 , depends on as @xmath17 . the index @xmath18 is then defined as @xmath19 . the observed shallow relation between cluster radius and mass ( e.g. @xmath20 ) ,
makes the value of the index @xmath21 similar to that found from observations and from simulations of clusters dissolving in tidal fields ( @xmath22 ) .
the constant of 2.0 gyr , which is the disruption time of a @xmath23 cluster in the solar neighbourhood , is about a factor of 3.5 shorter than found from earlier simulations of clusters dissolving under the combined effect of galactic tidal field and stellar evolution .
it is somewhat higher than the observationally determined value of 1.3 gyr .
it suggests , however , that the combined effect of tidal field and encounters with gmcs can explain the lack of old open clusters in the solar neighbourhood .
gmc encounters can also explain the ( very ) short disruption time that was observed for star clusters in the central region of m51 , since there @xmath14 is an order of magnitude higher than in the solar neighbourhood . |
the studies of multiparticle production can reveal high degrees of correlations among the spins of produced particles , i.e. particles would be produced with aligned spins .
chou and yang came to conclusion on existence of such correlations of particle spins on the grounds of their geometrical @xcite .
similar conclusion on spin correlations at very high energies has been made later on the base of the reflective scattering mechanism @xcite .
this mechanism will be discussed later in the context of the top quark priduction at the lhc . at the lhc
colliding proton beams are not polarized and the only way to perform spin - dependent mesurements is to determine the spin directions of final particles via their decay products kinematics .
one such possibility to detect global spin correlations @xcite is related to @xmath0-hyperons production and their spins measurements .
another possibility , almost exlusively appropriate to the lhc , is related to the top quark production .
this latter process has high statistics and also allows spin measurements through the restoration of the decay products kinematics . at the lhc ,
the standard model ( sm ) dominating mechanism of the top quark pairs production is due to gluon fusion .
spin correlations in the top - antitop production in the sm have been calculated in @xcite .
the main interest in their studies is related to sensitivity of those spin correlations to the presence of the extra dimensions @xcite . in present note
we briefly report the known results on these correlations and discuss the possible effects due to the reflective scattering mechanism .
we estimate the size and sign of these effects for the lhc .
as it was noted , in perturbative qcd the top - antitop pairs are produced at the lhc mainly in the gluon fusion processes @xcite , @xmath1 , and top decays prior to the hadronization .
this feature allows one to determine top quark spin direction by measuring corresponding angles of the decay products .
the decay modes in the sm consist of leptonic and hadronic ones : @xmath2 the direction of the top quark spin can be reconstructed via its decay products kinematics .
the differential decay rates can be written in the top quark rest frame in the form : @xmath3 where @xmath4 is the partial decay width of the particular channel , @xmath5 is the top - spin analysing power and @xmath6 is the angle between the top - spin axis and direction of momemtum of the particle @xmath7 .
the values of the analysing power for different particles have been calculated in the standard model ( cf .
e.g. @xcite ) .
the spin correlation parameter defined as @xmath8 can be extracted from the double differential angular distribution of the respective decay products with angles @xmath9 and @xmath10 : @xmath11/4.\ ] ] in eq .
( [ corr ] ) the @xmath12 are the cross - sections of the top - antitop pair production with respective spin directions . in the sm spin correlation parameter has been calculated for the lhc at the lowest order in @xmath13 and the values @xmath14 have been obtained .
the details can be found in @xcite , it should be noted that numerical values of spin correlations depend on the particular parametrizations of the parton distribution functions . besides
, the value of the top - quark spin correlations can be affected by several mechanisms considered in the next sections .
the possibility of the presence extra spacial dimensions in addition to the usual 3 + 1 dimensional manifold has been considered in @xcite .
the main purpose for the introduction of the new dimensions was solution of the hierarchy problem , i.e. huge difference in scales of electroweak and gravitational interactions .
the 4-dimensional planck scale @xmath15 is connected with the fundamental scale in @xmath16 dimensiones @xmath17 and radius of the compactified @xmath18 extra dimensiones @xmath19 by the following relation @xcite : @xmath20 when the radius @xmath19 is large ( @xmath21 mm for @xmath22 ) the value of @xmath17 can be around 1 tev and it allows one to solve the gauge hierarchy problem . in another theory
called 5-dimensional warped geometry theory @xcite our space is a five - dimensional anti de sitter space and the elementary particles except for the graviton are localized on a ( 3 + 1)-dimensional brane or branes .
we will not discuss the difference in these approaches , it is important to note that in the effective theory graviton propogating in the bulk can be expressed as an infinite tower of kaluza - klein ( kk ) spin-2 gravitons .
thus , in the 4-dimensional effective theory kk - gravitons interact with the sm fields .
this interactions change the total cross - section of the top quark production ( increase it by factor 2 for the values of @xmath17 around 0.5 - 1 @xmath23 ) .
it also leads to significant changes in the spin correlation parameter decreasing it from 0.3 in the sm to zero or even negative value in this region of @xmath17 .
this is due to the spin-2 nature of the kk gravitons . for the other values of @xmath17
the role of interactions with kk - gravitons is minimal @xcite .
there is another global dynamical mechanism which should be taken into consideration .
it is related to the reflective scattering .
the idea that reflective scattering can lead to the global spin correlations of the final particles has been considered in @xcite . in this section
we apply qualitative conclusions of the above paper to the top - quark production and estimate numerical value of the respective spin correlation parameter .
the notion of reflective scattering is related to appearance of the phase factor @xmath24 in the elastic scattering @xmath25 matrix element @xmath26 .
we discuss here for simplicity the case of pure imaginary scattering amplitude .
the negative sign of @xmath26 is a manifestation of the fact that the elastic scattering amplitude @xmath27 is greater than @xmath28 , note that @xmath29 .
reflective scattering is a result of unitarity saturation , when partial amplitude tends to unity at @xmath30 .
such saturation realised in the @xmath31matrix or rational form of the amplitude unitarization at high energies . thus ,
as it was noted already , the saturation of unitarity is characterized by the fact that beyond some threshold energy value the elastic scattering matrix element @xmath26 at @xmath32 becomes negative ( i.e. @xmath33 at @xmath34 ) and the inelastic overlap function @xmath35 starts to develop a peripheral impact parameter dependence since @xmath36 since @xmath37 . at the lhc energies this peripheral profile with a maximum at @xmath38 , where @xmath39 is the radius of reflective scattering determined by the relation @xmath40 , should become quite noticeable .
the usual exponential ( eikonal ) form of unitarization does not lead to such a dependence on the impact parameter .
the difference in the impact parameter dependencies is illustrated in fig .
the region around the values of the impact parameter @xmath38 has the highest probability of the multiparticle production given by the relation @xmath41 .
mechanism of reflective scattering leads to suppression of particle production at small impact parameters and the main contribution to the integral multiplicity @xmath42 comes from the region of @xmath43 .
thus , due to peripheral impact parameter dependence of the inelastic overlap function the secondary particles will be mainly produced at the impact parameters @xmath43 and this will lead to imbalance between orbital angular momentum in the initial and final states since the most secondary produced particles would carry large orbital angular momentum . to compensate this imbalance in the orbital momentum the spins of the secondary particles should be correlated , i.e. aligned .
such correlations are expected to appear at the energies where the reflective scattering occures .
= 100 mm such considerations are appropriate for all inelastically produced particles , i.e. the spin correlations have a global nature . despite that , it is difficult to determine spin directions of the most final particles . in this sense ,
the top quark is one of exclusions ( another one is @xmath0-hyperon productions for example ) , and as it was already noted , its spin can be directly measured through its decay . on the other side , production of top - quark pairs
is an inelastic process and it should be affected by unitarity constraints as well .
the imbalance in the orbital momentum will lead to spin correlations in the top - antitop production at the lhc where the reflective scattering should take place .
thus on the grounds of unitarity saturation , the value of spin correlation parameter @xmath44 should be proportinal to the relative imbalance of the orbital angular momentum and to estimate it we make the assumption and just equate these two quantities , i.e. @xmath45 where @xmath46 is the orbital angular momentum of the initial state in the region of reflective scattering and @xmath47 is the spin correlation parameter arising as a result of the reflective scattering . as it was discussed earlier ,
the spin correlation parameter has the same value for all other particles , but spin of @xmath48-quark can be determined due to its decay . the negative sign in eq .
( [ correlp ] ) appears due to the fact that imbalance in orbital angular momentum is compensated by the spin correlations of the final particles in order to provide conservation of the total angular momentum .
the numerical value of this spin correlation parameter can be calculated using inelastic overlap function ( cf .
1 ) @xmath49bdb}{\int_0^{r(s)}bdb}.\ ] ] since @xmath50 and @xmath51 we will have the following relation valid in the case when reflective scattering is included ( i.e. when saturation of unitarity is assumed at asymptotic energies ) : @xmath52 to obtain the numerical results specific models and parameter values should be used .
we apply the model developed in @xcite .
it is based on the rational form unitarization where @xmath26 has the form @xcite @xmath53 the explicit form of @xmath54 and values of the model parameters are described e.g. in @xcite .
the numerical values for the spin correlation parameter can be calculated according to eq .
( [ correlpn1 ] ) : @xmath55 thus , the reflective scattering mechanism at the lhc energies can significantly affect the expected values of the spin correlation parameters , e.g. it can compensate small values of spin correlation parameter predicted due to possible presence of the extra dimensions providing not accounted significant background .
there is a way to separate contribution to the spin correlation parameter coming from sm model gluon - gluon fusion , contribution of the virtual kk - gravitons and reflective scattering .
reflective scattering has a global nature and will lead e.g. to the value @xmath56 at 7 tev while former processes would provide zero value for this parameter due to independence of the double - parton scattering .
inderectly , this can be tested by measuring @xmath57 for example .
it should be noted that spin correlation parameter in the top - antitop production in the helicity basis has already been measured at the lhc by atlas @xcite ( with the value @xmath58 ) .
this number allows one to make a conclusion on the agreement with the sm predictions .
however , one also can speculate on the possibility of the presence of the another mechanism , namely the presence of contribution of the kk - gravitons which is corrected by the reflective scattering effects .
we discuss here the spin correlation parameter in the top quark production at the lhc .
this observable is sensitive to the effects of the extra dimensions and was considered in the literature as a clear signal of the kk - gravitons existence on the brane . at the same time there is an expectation of additional contribution into the spin correlations in top quark production at the lhc energies resulting from the possible existence of the prominent reflective scattering mechanism
. this can provide significant background for the detection of the new physics effects associated with the extra dimensions and should be taken into account under the correct interpretations of the experimental results .
the effects of reflective scattering can also be verified by the spin correlation parameter measurements in the @xmath0-hyperon production as it was discussed in @xcite .
we would like to thank r. chierici , v. kachanov , d. konstantinov , v. petrov , r. tenchini and a. uzunian for the discussions of the results and comments .
chou , c.n .
yang , int . journ . of mod .
phys . a * 2 , 1727 ( 1987 ) .
troshin , phys .
b * 597 , 391 ( 2004 ) g. mahlon , s.j .
parke , phys .
d * 53 , 4886 ( 1996 ) .
m. arai , n. okada , k. smolek , v. simak , phys .
d * 70 , 115015 ( 2004 ) .
m. arkani - hamed , s. dimopoulos , g. dvali , phys .
b * 436 , 263 ( 1998 ) .
l. randall , r. sundrum , phys .
lett . * 83 , 4690 ( 1999 ) .
troshin , n.e .
tyurin , phys .
d * 49 , 4427 ( 1994 ) s.m .
troshin , n.e .
tyurin , int . j. mod
. phys . a * 22 , 4437(2007 ) .
troshin , n.e .
tyurin , mod .
lett . a * 30 , 169 ( 2008 ) .
g. aad et al .
( atlas collaboration ) , phys .
* 108 , 212001 ( 2012 ) . * * * * * * * * * * | the specific effects of the reflective scattering ( antishadowing ) related to the spin correlations in the top quark production at the lhc are considered .
account for those effects is important in the searches for the signatures of the extra dimensiones .
it is shown that significant spin correlations could arise at the lhc energies due to reflective scattering and those can affect the signals of the extra dimensions . |
deep inelastic scattering ( dis ) at hera is a copious source of multi - jet events .
typical two - jet cross sections are in the 100 pb to few nb range and thus provide sufficiently high statistics for precision qcd tests @xcite .
clearly , next - to - leading order ( nlo ) qcd corrections are mandatory on the theoretical side for such tests .
full nlo corrections for one and two - jet production cross sections and distributions are now available and implemented in the @xmath5 jets event generator mepjet , which allows to analyze arbitrary jet definition schemes and general cuts in terms of parton 4-momenta@xcite . a variety of topics can be studied with these tools .
\1 ) the determination of @xmath0 over a range of scales @xmath1 from dijet production at hera : the dijet cross section is proportional to @xmath0 at leading order ( lo ) , thus suggesting a direct measurement of the strong coupling constant .
however , the lo calculation leaves the renormalization scale @xmath1 undetermined .
the nlo corrections substantially reduce the renormalization and factorization scale dependencies which are present in the lo calculations and thus reliable cross section predictions in terms of @xmath6 ( for a given set of parton distributions ) are made possible .
clearly , a careful study of the choice of scale in the dijet cross section is needed in order to extract a reliable value for @xmath7 .
we will present some arguments and studies of the scale dependence of the nlo dijet cross section which suggest that the average @xmath2 of the jets in the breit frame provides the `` natural '' scale for multi jet production in dis . here
, @xmath8 is defined by @xmath9 , where the subscripts @xmath10 and @xmath11 denote the jet and proton , respectively ( all quantities are defined in the breit frame ) .
2 ) the measurement of the gluon density in the proton ( via @xmath12 ) : the gluon density can only be indirectly constrained by an analysis of the structure function @xmath13 at hera @xcite .
the boson gluon fusion subprocess dominates the two jet cross section at low @xmath4 and allows for a more direct measurement of the gluon density in this regime .
a first lo experimental analysis has been presented in @xcite .
nlo corrections reduce the factorization scale dependence in the lo calculation ( due to the initital state collinear factorization , which introduces a mixture of the quark and gluon densities according to the dglap ( altarelli - parisi ) evolution ) and thus reliable cross section predictions in terms of the scale dependent parton distributions are made possible . some implications for the determination of the gluon density
have been discussed in @xcite .
\3 ) associated forward jet production in the low @xmath4 regime as a signal of bfkl dynamics : bfkl evolution @xcite leads to a larger cross section for events with a measured forward jet ( in the proton direction ) with transverse momentum @xmath14 close to @xmath15 than the dglap @xcite evolution .
clearly , next - to - leading order qcd corrections for fixed order qcd , with altarelli - parisi ( dglap ) evolution , are mandatory on the theoretical side in order to establish a signal for bfkl evolution in the data .
we discuss these corrections in section 3 .
the importance of higher order corrections and recombination scheme dependencies of the two jet cross sections for four different jet algorithms ( cone , @xmath16 @xcite , jade , w ) was already discussed in @xcite . while the higher order corrections and recombination scheme dependencies in the cone and @xmath16 schemes are small , very large corrections appear in the @xmath17-scheme .
we conclude from these studies that the cone and @xmath16 schemes appear better suited for precision qcd tests and will concentrate only on those schemes in the following .
the goal of a versatile nlo calculation is to allow for an easy implementation of an arbitrary jet algorithm and to impose any kinematical resolution and acceptance cuts on the final state particles .
this is best achieved by performing all hard phase space integrals numerically , with a monte carlo integration technique .
this approach also allows an investigation of the recombination scheme dependence of the nlo jet cross sections . for dijet production at hera
such a nlo monte carlo program is mepjet@xcite .
the calculation is based on the phase space slicing method and on the technique of universal `` crossing functions '' @xcite .
more details are described in @xcite and we do not repeat them here .
this technique can also be extended to the case of massive quark production @xcite .
an alternative technique for the calculation of nlo corrections in jet physics is the `` subtraction '' method @xcite .
jet production in dis is a multi - scale problem and it is not a priori clear at which scale @xmath18 is probed . in the following we will discuss the choice of the renormalization and factorization scale for dijet production at hera . as mentioned before ,
the nlo corrections are expected to reduce the scale dependencies in the lo predictions provided the scale is of the order of the typical hardness of the partonic process .
the following studies are done for the cone algorithm ( which is defined in the laboratory frame ) and the distance @xmath19 between two partons decides whether they should be recombined to a single jet . here
the variables are the pseudo - rapidity @xmath20 and the azimuthal angle @xmath21 .
we recombine partons with @xmath22 .
furthermore , a cut on the jet transverse momenta of @xmath23 gev in the lab is imposed .
we employ the two loop formula for the strong coupling constant with a value for @xmath24 consistent with the value from the parton distribution functions .
the value of @xmath18 is matched at the thresholds @xmath25 and the number of flavors is fixed to @xmath26 throughout , _
i.e. _ gluons are allowed to split into five flavors of massless quarks .
a running qed fine structure constant @xmath27 is used .
in addition the following set of kinematical cuts is imposed on the initial virtual photon and on the final state electron and jets : we require 40 gev@xmath28 gev@xmath29 , @xmath30 , an energy cut of @xmath31 gev on the scattered electron , and a cut on the pseudo - rapidity @xmath32 of the scattered lepton and jets of @xmath33 .
first studies of the scale dependence of the dijet cross section in the cone scheme are presented in @xcite .
we have considered scales related to the scalar sum of the parton transverse momenta in the breit frame , @xmath34 , and the virtuality @xmath35 of the incident photon . in the following we will also consider scales related to @xmath36 . here
, @xmath37 is defined by @xmath38 , where the subscripts @xmath39 and @xmath11 denote the final parton ( or jet ) and proton , respectively ( all quantities are determined in the breit frame )
. the breit frame is characterized by the vanishing energy component of the momentum of the exchanged photon .
both the photon momentum @xmath40 and the proton momentum @xmath41 are chosen along the @xmath42-direction .
@xmath4 is the standard bjorken scaling variable . in the parton model ,
the incoming quark with momentum @xmath43 collides elastically with the virtual boson and is scattered in the opposite direction ( the current hemisphere ) with momentum @xmath44 ; therefore , @xmath45 in the limit of the quark parton model , whereas @xmath46 .
the kinematics for dijet production is more complex : the momentum fraction @xmath20 of the incoming parton must be larger than @xmath4 since @xmath47 and , in general , the jets have a nonvanishing transverse momentum with respect to the @xmath48-proton direction . at lo ,
i.e. for massless jets , the relation between @xmath49 and @xmath50 reads : @xmath51 obviously , @xmath52 , and one can also show that @xmath53 .
thus , @xmath54 is approximately given by the harder of the two scales @xmath15 and @xmath55 @xcite .
for large dijet invariant masses ( i.e. for `` true '' two jet kinematics ) one has @xmath56 , the dijet system will be strongly boosted in the proton direction and @xmath57 , as long as very small scattering angles in the center of mass frame are avoided . for @xmath58 or smaller , on the other hand , @xmath59 and , typically , both are considerably larger than @xmath60 ( in particular for @xmath61 , which corresponds to the parton model limit ) . thus @xmath54 smoothly interpolates between the correct limiting scale choices , it approaches @xmath15 in the parton limit and it corresponds to the jet transverse momentum when the photon virtuality becomes negligible .
it appears to be the `` natural '' scale for multi jet production in dis . in @xcite
we found that the scale dependence of the dijet cross section does not markedly improve in nlo for @xmath62 .
this is also shown in fig . 1 ( dotted curves ) where the dependence of the two - jet cross section on the scale factor @xmath63 is shown .
we used the parton distribution functions set mrs d-@xmath64 @xcite and employed the two loop formula for the strong coupling constant both in the lo and nlo curves . for scales related to @xmath34 the uncertainty from the variation of the renormalization and factorization scale
is markedly reduced compared to the lo predictions ( dashed curves in fig . 1 ) .
here @xmath63 is defined via @xmath65 the resulting @xmath63 dependence for @xmath66 is shown as the solid lines in fig . 1 .
in this case , the nlo two - jet cross section is essentially independent on @xmath63 for @xmath67 .
hence , the theoretical uncertainties due to the scale variation are very small suggesting a precise determination of @xmath68 for different @xmath3 bins , where @xmath69 ( 7,7 ) ( 7,7 ) ( 7,7 ) ( 7,7 ) fig .
2 shows the @xmath3 distribution for the nlo exclusive dijet cross section .
we used the parton distribution functions set grv @xcite and @xmath70 .
in addition to the cuts imposed in fig . 1 , we require @xmath71 gev for each jet and @xmath72 gev .
we have divided the nlo cross section in the following three @xmath3 bins .
+ bin 1 : 5 gev @xmath73 10 gev + bin 2 : 10 gev @xmath74 15 gev + bin 3 : 15 gev @xmath75 .
+ table 1 shows the corresponding nlo cross sections for these three bins for two different sets of parton distributions and for different values for the scale factor @xmath63 .
the theoretical uncertainties for the nlo dijet cross section from these numbers are very small , in particular for the first two bins .
lccc & & & + + @xmath76 & 497 & 320 & 146 + @xmath77 & 504 & 306 & 134 + @xmath78 & 488 & 322 & 151 + @xmath76 & 487 & 322 & 148 + fig .
3 and 4 show the @xmath79 and @xmath35 distribution for the nlo exclusive dijet cross section for these three bins . whereas the @xmath79 and @xmath35 distributions are fairly similar to the @xmath3 distribution for the lowest bin , large differences are found for the other bins .
this partly explains the rather different scale dependence observed in fig.1 .
deep inelastic scattering with a measured forward jet with relatively large momentum fraction @xmath80 ( in the proton direction ) and @xmath81 is expected to provide sensitive information about the bfkl dynamics at low @xmath4 @xcite . in this region
there is not much phase space for dglap evolution with transverse momentum ordering , whereas large effects are expected for bfkl evolution in @xmath4 . in particular ,
bfkl evolution is expected to substantially enhance cross sections in the region @xmath82 @xcite . in order to extract information on the @xmath83 bfkl evolution
, one needs to show that cross section results based on fixed order qcd with dglap evolution are not sufficient to describe the data .
clearly , next - to - leading order qcd corrections to the dglap predictions are needed to make this comparison between experiment and theory . in table
[ table2 ] we show numerical results for the multi jet cross sections with ( or without ) a forward jet .
the lo ( nlo ) results are based on the lo ( nlo ) parton distributions from grv @xcite together with the one - loop ( two - loop ) formula for the strong coupling constant .
kinematical cuts are imposed to closely model the h1 event selection@xcite .
more specifically , we require @xmath84gev@xmath29 , @xmath85 , @xmath86 , an energy cut of @xmath87 gev on the scattered electron , and a cut on the pseudo - rapidity @xmath32 of the scattered lepton of @xmath88 ( corresponding to @xmath89 ) .
jets are defined in the cone scheme ( in the laboratory frame ) with @xmath90 and @xmath91 .
we require a forward jet with @xmath92 , @xmath93 gev , @xmath94 , and a cut on the pseudo - rapidity of @xmath95 ( corresponding to @xmath96 ) .
in addition all jets must have transverse momenta of at least 4 gev in the lab frame and 2 gev in the breit frame .
lcc & & + & & + + & 0 pb & 9026 pb + & 19.3 pb & 2219 pb + & 68 pb & 2604 pb + & 30.1 pb & 450 pb + the cross sections of table [ table2 ] demonstrate first of all that the requirement of a forward jet with large longitudinal momentum fraction ( @xmath97 ) and restricted transverse momentum ( @xmath94 ) severely restricts the available phase space , in particular for low jet multiplicities .
the 1-jet exclusive cross section vanishes at lo , due to the contradicting @xmath85 and @xmath97 requirements . for @xmath82 ,
a high invariant mass hadronic system must be produced by the photon - parton collision and this condition translates into @xmath98 where @xmath99 and @xmath100 are the transverse mass and rapidity of the partonic recoil system , respectively .
thus a recoil system with substantial transverse momentum and/or invariant mass must be produced and this condition favors recoil systems composed out of at least two additional energetic partons . as a result one finds very large fixed order perturbative qcd corrections ( compare 2 jet lo and nlo results with a forward jet in table [ table2 ] ) .
in addition , the lo @xmath101 3-jet cross section is larger than the lo @xmath102 2-jet cross section . thus , the forward jet cross sections in table [ table2 ] are dominated by the @xmath101 matrix elements .
the effects of bfkl evolution must be seen and isolated on top of these fixed order qcd effects .
we will analyze these effects in a subsequent publication .
the calculation of nlo perturbative qcd corrections has received an enormous boost with the advent of full nlo monte carlo programs @xcite . for dijet production at hera
the nlo monte carlo program mepjet @xcite allows to study jet cross sections for arbitrary jet algorithms . internal jet structure , parton / hadron recombination effects , and the effects of arbitrary acceptance cuts
can now be simulated at the full @xmath103 level .
we found large nlo effects for some jet definition schemes ( in particular the @xmath17-scheme ) and cone and @xmath16 schemes appear better suited for precision qcd tests .
the extraction of gluon distribution functions is now supported by a fully versatile nlo program .
preliminary studies show that large nlo corrections are present in the bjorken @xmath4 distribution for dijet events , while these effects are mitigated in the reconstructed feynman @xmath4 ( @xmath104 ) distribution , thus aiding the reliable extraction of @xmath105 . for the study of bfkl evolution by considering events with a forward `` mueller''-jet very large qcd corrections
are found at @xmath103 .
these fixed order effects form an important background to the observation of bfkl evolution at hera . they can now be studied systematically and for arbitrary jet algorithms .
this research was supported by the university of wisconsin research committee with funds granted by the wisconsin alumni research foundation and by the u. s. department of energy under grant no .
de - fg02 - 95er40896 .
the work of e. m. was supported in part by dfg contract ku 502/5 - 1 .
zeus collaboration , m. derrick et al . , z. phys . *
c65 * ( 1995 ) 379 ; + _ ibid .
b345 * ( 1995 ) 576 ; + h1 collaboration , t. ahmed et .
al . , nucl .
b439 * ( 1995 ) 471 ; + _ ibid .
* b354 * ( 1995 ) 494 . | next - to - leading order corrections to jet cross sections in deep inelastic scattering at hera are studied .
the predicted jet rates allow for a precise determination of @xmath0 at hera over a wide range of @xmath1 .
we argue , that the `` natural '' renormalization and factorization scale is set by the average @xmath2 of the jets in the breit frame and suggest to divide the data in corresponding @xmath3 intervals . some implications for the determination of the gluon density and the associated forward jet production in the low @xmath4 regime at hera
are briefly discussed . |
graphene a two - dimensional hexagonal lattice of carbon atoms was discovered about eight years ago @xcite and is now one of the most promising materials for future nanoelectronics .
the high application potential of this novel material is associated with some peculiarities of its electronic and phonon transport properties @xcite as well as with its outstanding mechanical @xcite and optical properties @xcite .
optoelectronic properties of graphene are also very promising for applications @xcite . moreover , owing to a very long spin relaxation time , which is expected due to a very weak spin - orbit interaction , graphene is also attractive for applications in spin electronics ( see , e.g. , ref .
however , to utilize the outstanding properties of graphene for spin - dependent transport , one needs to have a reliable method of controllable spin injection and spin manipulation .
the possibility of a relatively strong rashba spin - orbit coupling has been reported in refs . [ ] for graphene deposited on a ni ( or ni / au ) substrate .
such a strong spin - orbit coupling formally enables spin manipulation in graphene .
however , even in the absence of a substrate leading to strong spin - orbit coupling , experiments report spin relaxation time on the timescale of the order of or less than one nanosecond @xcite , which makes applications of graphene in spin electronics rather difficult . in all the experiments aimed at the measurements of spin relaxation time
, spins are injected _ en masse _ from a ferromagnetic contact giving rise to some charge / spin density distribution , which influences its subsequent dynamics .
several theoretical approaches ( see for example refs . [ ] ) have been proposed to describe spin relaxation . however , most of them demonstrated spin relaxation time much longer than that observed experimentally .
there are several experimental techniques which can be used to manipulate and control electron spin in graphene .
for example , spin current and spin density in graphene nanodisks can be manipulated by varying length of the corresponding zigzag edge @xcite .
quantum pumping of dirac fermions and spin current in a monolayer graphene in perpendicular magnetic field , with the gate voltage as a control parameter , has been proposed in ref .
furthermore , the method of spin current generation in a monolayer graphene through adiabatic quantum pumping by two oscillating in time potentials has been described in refs . [ ] and for the bilayer graphene in ref .
it is well - known that spin - orbit coupling can lead to a direct optical spin injection - the technique extensively used in the physics of semiconductors @xcite . for graphene , the spin - orbit coupling influences the optical response in the infrared frequency range @xcite . in this paper
we consider the infrared optical spin injection in a single - layer graphene by linearly polarized light .
the graphene is assumed to be deposited on a substrate which leads to the rashba spin - orbit coupling . due to this interaction , the electronic spectrum of graphene near the dirac
points splits into four bands with parabolic dependence on the electron momentum at small wave vectors and linear dispersion at large wave vectors@xcite .
splitting of the subbands is determined by the spin - orbit coupling strength .
we show that optical spin injection becomes allowed in the presence of an external magnetic field , and the injection efficiency is of the order of the ratio of the zeeman splitting and the spin - orbit coupling matrix element . by modifying the infrared light frequency , one can change the absorption region in the momentum space , and thus control the spin injection . in sec . 2 we derive some general formula for optical spin injection efficiency in graphene
numerical results on spin injection rate are presented and described in sec .
summary and final conclusions are in sec .
we assume an external magnetic field @xmath0 oriented in the graphene plane .
hamiltonian describing the low energy electron excitations near the dirac point @xmath1 in graphene with the rashba spin - orbit interaction takes then the form @xcite @xmath2 where @xmath3 , @xmath4 with @xmath5 being the coupling constant of rashba spin - orbit interaction,@xcite and @xmath6 is the land factor . the matrices @xmath7 and @xmath8 are the pauli matrices in the sublattice and spin space , respectively . the third term of the above hamiltonian stands for the rashba spin - orbit coupling induced by the substrate .
note , the first term is diagonal in the spin space and for abbreviation the corresponding unit matrix is not written explicitly .
similarly , the second term is diagonal in the sublattice space and the corresponding unit matrix is not written explicitly , too .
electronic spectrum corresponding to the hamiltonian ( [ 1 ] ) consists of four energy bands . in the limit of weak magnetic field ,
@xmath9 , this spectrum is described by the formulas @xmath10 with all possible combination of the @xmath11 and @xmath12 signs , and the index @xmath13 labeling the bands in the order of increasing energy ( see fig . 1 ) . as in the usual two - dimensional electron gas with rashba spin - orbit interaction ,
the expectation value of the spin @xmath14-component in eigenstates of hamiltonian ( 1 ) for @xmath15 is equal to zero .
however , unlike to the two - dimensional electron gas , the expectation value of the in - plane spin depends on the wave vector and is relatively small for low - energy electron states .
indeed , the expectation value , @xmath16 , of electron spin in the state @xmath17 for @xmath15 is given by @xcite @xmath18 where @xmath19 is the band index , @xmath20 for @xmath21 and @xmath22 for @xmath23 ( cf .
thus , when @xmath24 , the expectation value of spin is small , @xmath25 . moreover , the spins are perpendicular to the wave vectors , similarly as in two - dimensional electron gas .
we note that the upper index ( 0 ) at the eigenfunctions and eigenenergies indicates they are for @xmath15 . in the following ,
we take the in - plane magnetic field @xmath0 along the @xmath26-axis and assume it is rather weak , @xmath27 .
the former assumptions justifies the absence of landau quantization , while the latter condition assures that the band dispersion is only weakly perturbed by the static magnetic field , and the resulting spin injection is linear in the applied magnetic field @xmath28 .
the four - band structure is presented in fig .
[ fig1 ] , which also shows the assumed position of the fermi level @xmath29 . limit .
arrows show the @xmath30-dependent orientation of electron spins in the eigenstates of the hamiltonian . here
we consider only the vicinity of the @xmath1 dirac point , taking into account that the other valley gives exactly the same result for light absorption and spin injection . ]
taking into account the first term of eq .
( 1 ) , hamiltonian describing interaction of graphene with an external periodic electromagnetic field @xmath31 can be written as @xmath32 this periodic perturbation leads to electron transitions between the bands shown in fig . 1 .
the spin states of electrons involved in the transitions are then modified accordingly . without loss of generality ,
we assume in the following that the electromagnetic field is oriented along the @xmath33-axis , @xmath34 . the total absorption rate of photons can be calculated as the sum of all allowed transitions , @xmath35 where @xmath36 corresponds to the absorption associated with the transitions of electrons from the subband @xmath13 to the subband @xmath37 , which can be calculated from the fermi golden rule as @xmath38.\end{aligned}\ ] ] here , @xmath39 and @xmath40 are the eigenfunctions and eigenvalues of the total hamiltonian , and @xmath41 is the corresponding fermi distribution function .
it is convenient to introduce an independent of the system parameter @xmath42 , defined as @xmath43 and rewrite eq .
as @xmath44 with the system - dependent functions @xmath45 . since @xmath46 in eq . is related to the incident flux @xmath47 of @xmath33-polarized photons by the formula @xmath48 , eq . can be presented in the form @xmath49 the ratio @xmath50 corresponds to the absorption coefficient of graphene without rashba spin - orbit coupling . @xcite in the limit of large frequency , @xmath51 , the absorption rate ( 8) is constant and does not depend on frequency , like in the case of graphene with zero rashba coupling , @xmath52 .
thus , @xmath53 can be considered as a ratio of absorption coefficients for graphene with rashba spin - orbit interaction and of graphene without rashba interaction . in other words , @xmath53 is the absorption coefficient normalized to that for graphene without rashba interaction .
now , let us define the spin injection rate for the @xmath54-th component of the spin density .
following eq . , we write @xmath55.\end{aligned}\ ] ] similarly to the case of absorption , we introduce the total spin injection rate as @xmath56 and write @xmath57 and @xmath58 .
thus , @xmath59 and @xmath60 can be considered as normalized to @xmath61 spin injection rates . before discussing numerical results based on the above formula ,
let us discuss briefly physical origin of the spin injection . in the absence of magnetic field , symmetry of the matrix elements and spin expectation values ( as shown in fig .
[ fig1 ] ) as well as the independence of energy @xmath40 on the momentum orientation lead to zero spin injection rate , as required by the time - reversal symmetry . in an in - plane magnetic field , in turn ,
each subband is shifted in energy , @xmath62 as a result , the lines of energy conservation , @xmath63 for the transitions changing the electron spin , such as @xmath64 and @xmath65 , are not simple circles anymore and acquire a distortion of the order of @xmath66 , where @xmath67 is the angle between @xmath68 and the @xmath26-axis .
in addition , the 4-component wave functions are modified in the first order perturbation as @xmath69 accordingly , expectation values of the spin components @xmath70 and of the interband matrix elements @xmath71 acquire first - order modification in the applied magnetic field , which results in a nonzero spin injection .
now we present some numerical results on the absorption of linearly polarized light and the associated spin injection . in our calculations
we assumed the temperature @xmath72 k. mev .
( b ) the total normalized absorption coefficient for two different values of the rashba spin - orbit parameter , as indicated . both figures are calculated for magnetic field @xmath73 t and for the chemical potential @xmath74 mev . ] let us begin with the normalized absorption coefficients presented in fig .
[ fig2 ] as a function of the frequency @xmath75 . figure [ fig2]a shows the normalized absorption coefficients for individual inter - band transitions , while fig .
[ fig2]b shows the total normalized absorption coefficient for two different values of the rashba parameter @xmath5 . in the latter case
, the thin solid line corresponds to the absorption coefficient in the absence of rashba coupling .
as one can see in fig .
[ fig2 ] , the spin - orbit coupling strongly modifies the absorption , in agreement with the results of ref .
the frequency threshold for the interband transitions is determined by the pauli blocking and also depends on the chemical potential @xmath76 and the spin - orbit splitting @xmath77 . in the case
considered here , @xmath78 , the transitions of highest frequency occur between the @xmath79 and @xmath80 bands and start at the @xmath1-point with zero matrix element . at high frequencies , @xmath81 ,
the total absorption coefficient approaches that for a pure single - layer graphene without spin - orbit coupling . in a magnetic field of @xmath73 t parallel to the @xmath26-axis and external electromagnetic field polarized along the @xmath33-axis .
the dashed green ( solid red ) line corresponds to the coupling constant @xmath82 mev ( @xmath83 mev ) .
all results are for the chemical potential @xmath74 mev .
( b ) spin injection efficiency for spin change per absorbed photon for the same parameters and injection geometry . ]
let us consider now numerical results on spin injection shown in fig .
[ fig3 ] for the magnetic field @xmath73 t , which corresponds to the zeeman splitting @xmath84 of approximately 0.6 mev .
we show there only the injection rate for the spin @xmath26-component , and for clarity we simplified there the notation by omitting the index @xmath54 , @xmath85 .
figure [ fig3]a shows the total normalized spin injection rate as a function of frequency for two different values of the rashba coupling parameter . the dependence of the spin injection rate on the light frequency is rather complicated due to several interband transitions involved and complex dependence of the matrix elements on the transition frequency . for the spin - conserving transitions between the states characterized by the same @xmath86 in eq .
, such as @xmath87 and @xmath88 , the main contribution to the spin injection comes from changes in the expectation values of @xmath89 , while for the other transitions all the changes in the system make comparable contributions .
note , the spin injected is opposite to the magnetic field . the spin injection efficiency can be defined as the average spin injected by a single photon .
this efficiency is given by @xmath90 , and is shown in fig .
[ fig3]b for the same rashba parameters as in fig .
as one can see in fig .
[ fig3]b , the efficiency can reach 0.2 per incident photon .
in general , the injection rate is of the order of @xmath91 , and can be manipulated by changing the photon frequency in the range of the order of @xmath92 . since the transitions are rather complicated , the ratio @xmath91 should be considered as an order - of - magnitude estimate only . similar spin injection , though considerably weaker , can be obtained for the electric field along the magnetic field , i.e. along the @xmath26-axis .
we have considered theoretically spin injection in single - layer graphene in the presence of rashba spin - orbit coupling and in - plane external magnetic field .
we have found that the spin injection is efficient at frequencies of the order of spin - orbit band splitting , with the efficiency being of the order of the ratio of zeeman and rashba splittings . for experimentally achievable parameters of the spin - orbit coupling and magnetic field , the injection efficiency can achieve 0.2 per absorbed photon .
this result shows that optical spin injection opens a way for a controllable and efficient method of spin density and spin current generation in graphene , not requiring presence of any ferromagnetic contact .
this work is partly supported by the national science center in poland as a research project in years 2011 2014 and grants nos .
dec-2011/01/n / st3/00394 and dec-2012/06/m / st3/00042 .
the work of eys was supported by the university of basque country upv / ehu under program ufi 11/55 , spanish mec ( fis2012 - 36673-c03 - 01 ) , and grupos consolidados upv / ehu del gobierno vasco ( it-472 - 10 ) .
t. y. yang , j. balakrishnan , f. volmer , a. avsar , m. jaiswal , j. samm , s. r. ali , a. pachoud , m. zeng , m. popinciuc , g. gntherodt , b. beschoten , and b. zyilmaz , phys .
lett . * 107 * , 047206 ( 2011 ) .
d. v. fedorov , m. gradhand , s. ostanin , i. v. maznichenko , a. ernst , j. fabian , and i. mertig , phys .
lett . * 110 * , 156602 ( 2013 ) , s. konschuh , m. gmitra , d. kochan , and j. fabian , phys .
b * 85 * , 115423 ( 2012 ) . | we calculate the efficiency of infrared optical spin injection in single - layer graphene with rashba spin - orbit coupling and for in - plane magnetic field . the injection rate in the photon frequency range corresponding to the rashba splitting is shown to be proportional to the ratio of the zeeman and rashba splittings . as a result , large spin polarization can be controllably achieved for experimentally available values of the spin - orbit coupling and in magnetic fields below 10 tesla . |
equation ( [ surface_int ] ) gives the surface integral that determines the 3-body elastic interaction among nano - column inclusions .
integration by parts combined with @xmath130 yields that it is symmetric with respect to the latter : @xmath131 . in the case that @xmath132 , it reduces to the angular integral @xmath133 where @xmath134 . here
, @xmath135 denotes the orientation of the vector @xmath136 . after making the change of variables @xmath137 , application of cauchy s theorem
yields that the integral vanishes : @xmath138 . in the case that @xmath139 and @xmath140 , the surface integral ( [ surface_int ] ) reduces to @xmath141 . \label{ijk}\ ] ] repeating the previous steps results in a closed - form expression with a large number of terms .
symbolic manipulation programs then help reduce these to the result ( [ 3-body ] ) .
a. goyal , s. kang , k.j .
leonard , p.m. martin , a.a .
gapud , m. varela , m. paranthaman , a.o .
ijaduola , e.d .
specht , j.r .
thompson , d.k .
christen , s.j .
pennycook and f.a list , supercond .
* 18 * , 1533 ( 2005 ) . | the results of a linear elasticity analysis yields that nano - rod inclusions aligned along the @xmath0 axis of a thin film of yba@xmath1cu@xmath2o@xmath3 , such as bazro@xmath2 and basno@xmath2 , squeeze that matrix by pure shear .
the sensitivity of the superconducting critical temperature in that material to the latter implies that the phase boundary separating the nano - rod inclusion from the superconductor acts as a collective pinning center for the vortex lattice that appears in external magnetic field . a dominant contribution to the in - field critical current can result .
the elasticity analysis also finds that the growth of nano - rod inclusions can be weakly metastable when the inclusion is softer than the matrix . _
introduction . _
the ongoing development of thin films of superconducting yba@xmath1cu@xmath2o@xmath3 ( ybco ) for wire technology has resulted in world - record high critical currents.@xcite in external magnetic field , the critical current is considerably enhanced by nano - rod inclusions that are aligned in parallel to the crystalline @xmath0 axis.@xcite@xmath4@xcite@xmath4@xcite@xmath4@xcite the enhancement is strongest at orientations of the magnetic field parallel to the @xmath0 axis .
understanding the fundamental physics behind this effect remains a challenge .
it is also unknown what drives the growth of nanorods in the first place in ybco films . in this paper , we provide insight into both of these questions by computing the strain field due to nano - rod inclusions that thread a ybco superconductor along the @xmath0 axis .
the lattice constant of inclusions that optimize the critical current is typically 8% larger than that of the ybco matrix in the @xmath5-@xmath6 plane . assuming a coherent phase boundary between the inclusion and a given epitaxial layer of ybco
, a linear elasticity analysis yields that the nanocolumn is compressed axially , while the ybco matrix is squeezed by pure shear about the nanocolumn .
the critical temperature in optimally doped ybco is known to couple strongly to pure shear in the @xmath5-@xmath6 plane.@xcite in applied magnetic field , we show theoretically how this experimental fact results in substantial collective pinning of the vortex lattice by the phase boundary separating the nano - column inclusion from the ybco matrix.@xcite@xmath4@xcite also , the elastic energy shows weak metastability at a high density of nanocolumns when the nanocolumn is soft compared to the ybco layer ( see fig . [ plot_energy ] ) .
we believe that this drives epitaxial growth of nano - rod inclusions in ybco films . _
two - dimensional elasticity theory .
_ we shall determine first the elastic strain and the elastic energy cost due to a single nano - rod inclusion that threads a film of ybco along the @xmath0 axis .
such nanorods are typically composed either of bazro@xmath2 ( bzo)@xcite@xmath4@xcite or of basno@xmath2 ( bso).@xcite both are cubic perovskites , with lattice constants ( @xmath7 ) that exceed that of the @xmath5-@xmath6 plane in ybco , @xmath8 , by 9% and by 7% respectively.@xcite@xmath4@xcite@xmath4@xcite let us temporarily ignore the effect of the lattice mismatch along the @xmath0 axis by considering only epitaxial layers that are far from any possible partial misfit dislocation , and that therefore present a coherent phase boundary between the inclusion and the ybco matrix .
such partial misfit dislocations are accompanied by stacking faults,@xcite a topic which will be discussed later in the concluding section .
the assumption of a coherent phase boundary is valid for a nano - rod inclusion of diameter less than the distance between possible misfit dislocations,@xcite @xmath9 .
bzo nanorods typically have a diameter of@xcite @xmath10 , which satisfies the bound @xmath11 .
bso nanocolumns , on the other hand , typically have a diameter of@xcite @xmath12 .
it exceeds @xmath13 , although not by much .
consider then a cylindrical nano - column inclusion that presents a coherent phase boundary with a given epitaxial layer of the ybco matrix .
unit cells match up one - to - one across the phase boundary in such case .
the ideal axial symmetry , assumed here for simplicity , implies a radial displacement field , @xmath14 .
we then have the boundary condition @xmath15 between the displacement fields of the nanocolumn ( in ) and of the ybco layer ( out ) at the phase boundary .
the in - plane lattice mismatch that it represents generates elastic strain in both the inclusion and in the ybco matrix .
the elastic energy due to a 2d strain field is given by the integral@xcite @xmath16\biggr\ } \label{e_2d}\ ] ] over the corresponding area ( prime ) , which is confined to @xmath17 for the nanocolumn and to @xmath18 for the ybco matrix .
here , @xmath19 and @xmath20 are the 2d bulk compression modulus and the 2d shear modulus , respectively .
a useful identity for the pure shear component above reads @xmath21 the strain tensor takes the form @xmath22 in the present axially symmetric case .
it combined with eq .
( [ identity ] ) results in the compact expression for the elastic energy , @xmath23 ^ 2 + { 1\over 2 } c_{\perp } [ r\ , d ( r^{-1 } u ) / d r]^2\}$ ] .
calculus of variations then yields a nano - column inclusion squeezed by pure compression and a surrounding ybco matrix squeezed by pure shear : @xmath24 with corresponding strain tensors @xmath25 the total elastic energy ( [ e_2d ] ) generated by the nano - column inclusion is then @xmath26 .
minimizing it with respect to the constants @xmath27 and @xmath28 while enforcing the boundary condition ( [ bc_1 ] ) yields optimal values @xmath29 and @xmath30 . here
@xmath31 .
these then yield an elastic energy cost @xmath32 for the nano - column inclusion , which has an equilibrium radius @xmath33 given by @xmath34 .
consider next a field of many cylindrical nano - column inclusions of radius @xmath35 centered at transverse locations @xmath36 .
suppose again that they all present a coherent phase boundary with a given epitaxial layer of the ybco matrix .
the displacement field is then a linear superposition of those generated by a single nano - column inclusion ( [ sol ] ) : @xmath37 % \quad{\rm for}\quad |{\bf r } - { \bf r}_i| < r_{\rm in } , \label{sol_in}\ ] ] inside the @xmath38 nanocolumn , and @xmath39 inside the ybco matrix .
the pure shear terms that have been added to the pure compression inside of a nanocolumn ( [ sol_in ] ) are required by the boundary condition ( [ bc_1 ] ) .
observe now , by eq .
( [ strain ] ) , that @xmath40 .
inspection of the elastic energy functional ( [ e_2d ] ) combined with the identity ( [ identity ] ) then yields that the above superpositions are stationary because @xmath41 , @xmath42 and @xmath43 all vanish .
indeed , the elastic energy cost reduces to a sum of surface integrals around the phase boundaries of the form @xmath44 + \sum_i \sum_{j , k}^{\prime } [ e_{i , j , k}({\rm in } ) + e_{i , j , k } ( { \rm out})]$ ] , where the indices @xmath45 and @xmath46 refer to the terms in the superpositions ( [ sol_in ] ) and ( [ sol_out ] ) , and where the index @xmath47 refers to the phase boundary .
the prime notation over the summation symbols indicates that @xmath48 .
each individual contribution @xmath49 is given by a surface integral around the circle @xmath50 of radius @xmath51 that is centered at @xmath52 : @xmath53 , with @xmath54 \cdot { \bf u}_{\perp } ( { \bf r } - { \bf r}_k ) .
\label{surface_int}\ ] ] here , @xmath55 and @xmath56 .
also , the measure @xmath57 on the circle @xmath50 points radially outward . substituting in the strain fields ( [ strain ] ) above yields ultimately that @xmath58 , and that @xmath59^{-2 } \label{3-body}\ ] ] for @xmath60 .
( see appendix . ) here , @xmath61 , and @xmath62 denotes the angle between the vectors @xmath63 and @xmath64 .
the 2d elastic energy then is composed of a sum of 1-body , 2-body ( @xmath65 ) and 3-body terms ( @xmath66 ) , @xmath67 , with the interaction energy given by @xmath68 b_0 ^ 2 r_{\rm out}^6 { \rm re}\ , [ r_{i , j } r_{i , k } e^{i\phi_{j , k } ( i ) } - r_{\rm out}^2]^{-2}. \label{interaction}\ ] ] notice that @xmath69 changes sign as a function of the relative rigidity between the nano - column inclusion and the ybco matrix .
the elastic energy will now be obtained by computing subsequent self - energy corrections to the 2-body interaction and to the 1-body line tension .
let s first fix the coordinate for the phase boundary above , @xmath52 , as well as one of the nanocolumn coordinates above , @xmath70 .
observe that the 3-body interaction ( [ interaction ] ) has zero angle average about the center @xmath52 over the remaining nanocolumn coordinate @xmath71 .
this is due simply to the fact that the contour integral @xmath72 around the unit circle , @xmath73 $ ] , vanishes for complex @xmath74 inside of that circle .
let s assume that each nanocolumn has a hard core of radius @xmath75 . at @xmath76 ,
we then obtain the estimate @xmath77 for the correction to the 2-body interaction on average over the bulk of the system .
here , @xmath78 denotes the density of nanocolumns .
the renormalized 2-body interaction that results is then @xmath79 .
next , assume an effective hard - core of radius @xmath80 for the nanocolumn at the coordinate @xmath70 that remains .
we thereby obtain the estimate @xmath81 for the net self - energy correction to the elastic energy of an isolated nano - column inclusion , with @xmath82 \cdot [ r_{\rm out}^4 / ( r_2^{\prime 2 } - r_{\rm out}^2 ) ] .
\label{r_2}\ ] ] this yields a total elastic energy density @xmath83 n_{\phi } e_{\rm 2d}^{(1 ) } \label{energy}\ ] ] as a function of the density of nanocolumns .
the above third - order polynomial is depicted by fig .
[ plot_energy ] .
it notably predicts weakly metastable epitaxial growth for relatively soft nanorods within the ybco matrix , such that @xmath84 .
this occurs at a density @xmath85^{1/2})/3\pi r_1 ^ 2 $ ] of nano - rod inclusions , at large effective crossections @xmath86 .
the equilibrium density of nano - rod inclusions therefore _ can not _ be dilute .
in particular , @xmath87 must lie somewhere between @xmath88 and @xmath89 .
inspection of eq .
( [ r_2 ] ) indicates that the former condition requires some degree of agglomeration among the nano - column inclusions : @xmath90 .
this may , however , be an artifact of the previous estimate for the 2-body self - energy correction , which is not accurate at @xmath91 .
last , the elastic energy cost per unit volume ( [ energy ] ) at meta - stable equilibrium is @xmath92 in the marginally stable limit at @xmath93 ( see fig . [ plot_energy ] ) .
the strong dependence that it shows on the bulk compression modulus of the inclusion affects growth dynamics .
this could be the root cause for the difference in length between bzo nanorods and bso nanocolumns in ybco.@xcite _ critical current by two - dimensional collective pinning .
_ we shall now determine the critical current of a thin film of superconducting ybco threaded by nano - rod inclusions along the crystalline @xmath0 axis and subject to external magnetic field aligned along the same axis .
recall that the critical temperature in an optimally doped ybco superconductor is primarily sensitive to _ shear _ strain in the @xmath5-@xmath6 plane.@xcite that fact coupled with the shear strain generated by a nano - column inclusion ( [ strain ] ) results in a potential - energy landscape for vortex lines that can collectively pin the vortex lattice .
in particular , the contribution of the vortex core to the vortex line tension is approximated by the fundamental energy scale per unit length @xmath94 , where @xmath95 denotes the london penetration depth . the temperature dependence shown by the vortex line tension
is therefore approximated by @xmath96 $ ] near the mean - field critical temperature @xmath97 .
the potential - energy landscape experienced by a vortex line then has a contribution @xmath98 , where @xmath99 is the true critical temperature , and where @xmath100 is the symmetric strain tensor ( [ strain ] ) .
it results in a @xmath101-wave potential - energy landscape about the nanocolumn for a vortex core , @xmath102 b_0 . \label{potential}\ ] ] with @xmath103 b_0 $ ] . here
the ratio between @xmath104 and @xmath105 is assumed to be constant .
a rigid vortex line therefore experiences a force field @xmath106 due to the strain generated by a single nano - column inclusion , where @xmath107 is the maximum force per unit length .
the above pinning / anti - pinning force ( [ force ] ) is long range .
the presence of an extended field of nanocolumns can cut the range off , however .
( see fig .
[ plot_landscape ] . ) such forces add within the present elastic approximation ( [ sol_out ] ) : @xmath108 .
the @xmath101-wave nature of each isolated force field ( [ force ] ) implies a null net force on average .
a characteristic fluctuation of the force over the ybco matrix remains : @xmath109 , where integration ( prime ) is restricted to the ybco matrix . matching @xmath110 with @xmath111 yields an effective range for each pinning / anti - pinning center @xmath112 the @xmath101-wave potential ( [ potential ] ) that acts on rigid vortex lines in the vicinity of the phase boundaries between the nano - column inclusions and
the ybco matrix has zero angle average .
it therefore can not pin down a vortex line in isolation .
previous work by one of the authors and maley@xcite implies that many of them collectively pin the abrikosov vortex lattice , however .
a hexatic bose glass state can exist at low temperature.@xcite it is a vortex lattice threaded by isolated lines of edge dislocations in parallel to the relatively weak correlated pinning / anti - pinning centers .
plastic creep of the vortex lattice associated with glide by such edge dislocations limits the critical current,@xmath113 which is given by @xmath114 . here
@xmath115 denotes the density of vortex lines pinned by the nanocolumns , @xmath116 is the elastic shear modulus of the pristine vortex lattice at a density @xmath117 of vortex lines,@xcite and @xmath6 denotes the magnitude of the burgers vector associated with the edge dislocations that thread the vortex lattice .
the @xmath101-wave nature of the pinning / anti - pinning center ( [ potential ] ) also implies that its occupation is purely random .
the density of vortex lines that they collectively pin is then equal to @xmath118 , where @xmath119 is the effective crossectional area of a pinning / anti - pinning center ( see fig .
[ plot_landscape ] ) .
the critical current density therefore obeys a _ pure _ inverse - square - root power law with magnetic field , @xmath120 . taking values of @xmath121 and @xmath122 for the strain derivatives of @xmath99 in optimally - doped ybco@xcite can result in a pinning efficiency , @xmath123 , of 93% at liquid nitrogen temperature !
_ discussion and conclusions .
_ we have found that the growth of nano - rod inclusions in ybco films is very likely driven by weak metastability shown by the elastic energy of epitaxial layers .
we also have pointed out how the sensitivity of the critical temperature in optimally - doped ybco to pure shear strain inside of the @xmath5-@xmath6 plane@xcite results in an effective collective pinning center for the abrikosov vortex lattice at the phase boundary between the nano - rod inclusion and the ybco matrix .
the lattice mismatch along the @xmath0 axis between the nano - rod inclusion and ybco has so far been neglected , however .
ybco has a unit cell that can be divided into a stack of three cubes along the @xmath0 axis , each with a lattice constant @xmath124 .
the strain that results at the phase boundary with a bzo nanorod or with a bso nanocolumn , both of which are cubic with lattice constants @xmath125 and @xmath126 , respectively , can be relieved by introducing partial misfit dislocations accompanied by stacking faults in the ybco matrix.@xcite the predicted spacing between such stacking faults , @xmath127^{-1}$ ] , is then equal to @xmath128 for bzo nanorods and to @xmath129 for bso nanocolumns ( cf .
ref .
5 ) . since their effect on the previous elasticity analysis can be accounted for by renormalized elastic moduli for the ybco matrix , we believe that that our conclusions remain unchanged in their presence . the authors thank george levin for discussions .
this work was supported in part by the us air force office of scientific research under grant no .
fa9550 - 06 - 1 - 0479 . |
the flow of a dc current in response to a slowly - varying time - dependent potential operating on an _ un - biased _ system is termed
adiabatic charge pumping " .
@xcite this phenomenon , first considered in ref . ,
has attracted recently much theoretical @xcite and experimental @xcite interest . in general , adiabatic pumping " occurs when the charge transferred across a boundary during a single period of a certain modulating potential is independent of the modulation frequency .
this process is adiabatic in the sense that the periodic potential varies very slowly in time , such that its frequency @xmath0 is smaller than any characteristic energy scale of the electrons .
eighteen years ago , thouless @xcite has shown , using the adiabatic approximation , that the ground state of an infinite one - dimensional ( 1d ) system of non - interacting electrons subject to a slowly moving periodic potential can support a dc current .
later theoretical investigations of quantum pumping in confined nanostructures have employed the result derived by brouwer , @xcite which gives the pumped charge in terms of the time - dependent scattering matrix related to the modulating potential .
the derivation presented by brouwer is based on the analysis of ref . ,
which , in turn , utilizes an expansion in the amplitude of the modulating potential in the context of time - dependent scattering theory .
nevertheless , it is generally accepted that the formalism of bttiker _
_ @xcite and the resulting brouwer formula are valid @xcite even for large amplitudes , as long as the adiabatic approximation holds
. this paper is devoted , among other issues , to the exploration of this point .
we consider a spatially - confined system of non - interacting electrons , connected by leads ( denoted by @xmath1 ) to electronic reservoirs which are kept at various chemical potentials , @xmath2 .
the system is also subject to a slowly - varying periodic potential .
we derive an expression for the instantaneous current flowing in this system , allowing for arbitrary strengths of the modulating potential , using time - dependent scattering states .
the formalism is based on an iterative solution of those states , in the adiabatic approximation , in which the temporal derivative of the scattering potential ( and the scattering states ) is the small parameter .
the formal derivation is summarized in sec .
ii , where we obtain the instantaneous current in the lowest - order adiabatic approximation .
this current is averaged over a single period of the modulating potential . in this way we obtain the effect of the chemical potential bias on the pumped charge on one hand , and the modification of the landauer formula caused by the modulating potential on the other .
we find a new expression for the charge pumped through an un - biased system [ see eq .
( [ intpump ] ) below ] , which is particularly useful in cases where the modulating potential operates on the entire nanostructure , @xcite and is spatially - dependent .
the appendix includes the derivation of the next order correction to the current .
we investigate the lowest - order expression for the current in sec .
iii , confining ourselves for simplicity to a system connected to two reservoirs .
our results there can be summarized as follow . in a two - terminal structure ,
the current averaged over a single period , @xmath3 , consists of two parts .
the first , denoted by @xmath4 , flows even when the system is un - biased , ( but is modified by the presence of the chemical potential difference ) . to the lowest order in the adiabatic approximation , that current reads @xmath5,\label{intpump}\end{aligned}\ ] ] where @xmath6 and @xmath7 denote the left and the right leads , respectively , and @xmath8 are the fermi distributions in the reservoirs connected to the left and right leads . in eq .
( [ intpump ] ) , @xmath9 is the instantaneous scattering state at time @xmath10 , excited by an incoming wave in channel @xmath11 of lead @xmath12 , and @xmath13 is the temporal derivative of the time - dependent scattering potential .
interestingly enough , the matrix element appearing in eq .
( [ intpump ] ) can be written in terms of the instantaneous scattering matrix , reproducing the brouwer formula , @xcite and vindicating its use for arbitrary amplitudes .
the second part of the current , denoted by @xmath14 , flows only when the system is biased , i.e. , @xmath15 .
when this current is integrated over a single period , @xmath3 , of the modulating potential , one obtains the landauer formula , modified by the modulated potential in two ways : ( i ) the transmission coefficient is the instantaneous one , @xmath16 , averaged over a single period of the time - dependent potential ; ( ii ) there appears a correction to the landauer expression , which is related to the temporal derivative of the modulating potential and the ensuing instantaneous scattering matrix . in the simplest case in which the two leads are single - channel ones , @xmath14 takes a particularly simple form , @xmath17\bigl ( f_{\ell}(e)-f_{r}(e)\bigr ) , \label{landauer}\end{aligned}\ ] ] in which @xmath18 is the instantaneous friedel phase ( i.e. , the transmission phase ) of the nanostructure . the first term in eq .
( [ landauer ] ) yields the landauer formula for the present case . the second term there is a correction , which is discussed in sec .
iii , using a simple example .
in the first part of this section we solve for the time - dependent scattering states iteratively , using the adiabatic approximation .
we then use those scattering states in the second part , to obtain the current .
the formalism presented below borrows from the derivations in refs . and , extended to include the effect of a time - dependent scattering potential .
we consider a ballistic nanostructure of arbitrary geometry , which consists of a nanostructure connected to several electronic reservoirs .
this system is described by the hamiltonian @xmath19 where the scattering potential @xmath20 is assumed to be confined in space , so that asymptotic behaviors of the scattering solutions can be defined un - ambiguously .
this confined region is attached to leads , numbered by the index @xmath1 , and each lead is connected to a reservoir having the chemical potential @xmath2 .
the hamiltonian @xmath21 consists of the kinetic energy .
we use the adiabatic approximation , requiring that the characteristic inverse time - constant , @xmath22 , which describes the time dependence of @xmath23 , is smaller than any characteristic energy scale of the electrons . for a simple oscillatory potential , @xmath24 . as in the usual scattering treatment
, we denote the incoming wave with energy @xmath25 in lead @xmath1 by @xmath26 , where @xmath27 is the transverse mode number .
this wave is a solution of the free hamiltonian , @xmath28 and is normalized such that it carries a unit flux .
the scattering solution of the full hamiltonian , excited by this incoming wave , can be written in the form @xmath29 the time dependence of the scattered wave function , @xmath30 , is expected to have the same characteristic time scale as @xmath23 .
for example , when the modulating potential is oscillating in time , @xmath31 contains all harmonics of the frequency @xmath0 . the scattering solution @xmath32 should satisfy the time - dependent schrdinger equation , @xmath33 inserting ( [ scatsol ] ) into ( [ sch ] ) , using eq .
( [ h0 ] ) , we find @xmath34 where @xmath35 is the instantaneous green function of the full hamiltonian , such that @xmath36 we now solve eq .
( [ chitild ] ) , using the adiabatic approximation : the temporal derivative appearing on the right - hand - side of that equation is regarded as a small correction ( of order @xmath22 ) , and the equation is solved iteratively .
the zeroth - order is just the instantaneous scattering solution , which we denote by @xmath37 , @xmath38 the instantaneous scattering state @xmath37 is the solution of the instantaneous schrdinger equation , with energy @xmath25 , @xmath39 turning back to eq .
( [ chitild ] ) , one finds that to first order in the time - derivative the scattering solution reads @xmath40 in which @xmath41 is the time - derivative of the instantaneous scattering state .
hence , to first order in the adiabatic approximation the time - dependent scattering states are given entirely in terms of the _ instantaneous _ solutions ( namely , @xmath42 and @xmath43 ) of the problem at hand .
one notes that the adiabatic solution ( [ sol ] ) of the scattering state is analogous to thouless @xcite solution for the ground - state wave function in his model . in the appendix
, we discuss the corrections to the lowest order adiabatic approximation . here
we outline the derivation of the current in the scattering states formalism , as developed , e.g. , in refs .
one writes the field operator of the electron , @xmath44 , in terms of the scattering states as @xmath45 in which @xmath46 destroys an electron incoming in channel @xmath27 of lead @xmath1 .
the thermal average of the latter operators is given by the fermi distributions of the various reservoirs , such that @xmath47 where @xmath48 is the fermi distribution in the reservoir connected to the @xmath1 lead .
with these definitions , the thermal average of the current density operator becomes @xmath49 where @xmath50 stands for the negative electron charge .
it is convenient to evaluate this quantity when @xmath51 approaches @xmath52 in lead @xmath12 ( which will be denoted by @xmath53 ) , and then to integrate the current density over the cross - section of that lead ( noting that the incoming and outgoing waves are normalized to carry a unit flux ) . in so doing , we may take advantage of the asymptotic properties of the instantaneous quantities @xmath37 and @xmath54 , as documented in refs . and
, @xmath55 here , @xmath56 is the outgoing wave in channel @xmath11 of lead @xmath12 and @xmath57 is the matrix element of the instantaneous scattering matrix . as a result , the current flowing into lead @xmath12 is given by @xmath58 \biggr ) , \label{gencur}\end{aligned}\ ] ] with @xmath59 the result ( [ gencur ] ) for the time - dependent current entering into lead @xmath12 of the nanostructure holds for a general biased system , whose various terminals have different chemical potentials ( as long as the time - dependence of the periodic potential is slow enough ) .
it is therefore interesting to consider charge conservation , using that result .
indeed , summing ( [ gencur ] ) over all leads , we obtain @xmath60 which shows that when the total current entering the system , @xmath61 , is integrated over a single period of the modulating potential , the result is zero , i.e. , the charge per period is conserved . in deriving the result ( [ chacon ] ) we have employed ( i ) the unitarity of the instantaneous scattering matrix , @xmath62 ; and ( ii ) the following property of the scattering matrix @xcite @xmath63 equation ( [ gencur ] ) can be considered as a generalization of the landauer formula , extended to include the effect of a time - dependent potential , in the adiabatic approximation .
the new ingredient is the quantity @xmath64 , eq .
( [ u ] ) .
this quantity can be expressed in terms of the temporal derivative of the scattering potential , @xmath65 to prove this , we take the temporal derivative of eq .
( [ chiteq ] ) , and use eq .
( [ green ] ) , to obtain @xmath66 we insert this expression into eq .
( [ u ] ) , and carry out one of the spatial integrations using @xmath67 which follows directly by differentiating eq .
( [ chiteq ] ) with respect to the energy , and using eq .
( [ green ] ) and the symmetry of the green function @xmath68 .
this produces the result ( [ uder ] ) .
let us now confine ourselves to a nanostructure connected to two terminals , with left ( @xmath69 ) and right ( @xmath7 ) leads .
then we can use eq .
( [ gencur ] ) to write the current entering the system from the left terminal in the form @xmath70.\end{aligned}\ ] ] an analogous expression holds for @xmath71 .
the net current flowing in the system during a single period of the modulating potential then consists of two parts , @xmath72 where the first , @xmath14 , flows only when the system is biased , whereas the second , @xmath4 , is established by the time - dependent potential ( though it is affected by the chemical potential difference , when the latter is applied ) . using eqs .
( [ sums ] ) and ( [ uder ] ) , the pumped current , @xmath4 , takes the form @xmath73 . \label{ipump}\end{aligned}\ ] ] for the biased current we find @xmath74.\label{ibias}\end{aligned}\ ] ] the pumped part of the current , @xmath4 , can be written in terms of the temporal derivatives of the instantaneous scattering matrix . in order to vindicate this statement
, we start from the asymptotic form for @xmath75 , eq .
( [ asymp ] ) , for the wave going from lead @xmath1 into lead @xmath12 .
noting that @xmath76 we conclude that @xmath77 differentiating this expression with respect to time and using eqs .
( [ chit ] ) and ( [ chitt ] ) , yields @xmath78 note that this expression is * not * a matrix element . to turn it into an expression involving matrix elements
, we use eq .
( [ sums ] ) , by which @xmath79 then , multiplying by @xmath80 and summing over @xmath1 and @xmath27 yields @xmath81 one notes that this identity makes the expression for @xmath82 , eq .
( [ ipump ] ) above , to be identical with the brouwer @xcite formula , in the case where the system is un - biased . in the simplest case where each of the leads is a single - channel one , the expression for the current takes a particularly simple form .
in this situation , the instantaneous scattering matrix becomes a @xmath83 matrix , which can be parametrized ( in the absence of a magnetic field ) as @xmath84.\label{smatrix}\end{aligned}\ ] ] here , @xmath16 and @xmath85 are the instantaneous transmission and reflection , respectively .
the reflection phase @xmath86 describes the asymmetry of the nanostructure .
a finite time - dependent reflection phase is a necessary ingredient to obtain the pumped current .
finally , @xmath18 is the transmission ( friedel ) phase . with the parametrization eq .
( [ smatrix ] ) one finds , using eqs .
( [ sdot ] ) and ( [ brouwer ] ) , @xmath87\nonumber\\ & + & \bigl ( f_{\ell}(e)+f_{r}(e)\bigr ) \frac{\partial}{\partial e}\bigl ( r^{t}\frac{d\alpha^{t}}{dt}\bigr ) \biggr ] .\end{aligned}\ ] ] the charge passing through the nanostructure during a single period of the potential , @xmath88 , is then @xmath89 here , @xmath90.\end{aligned}\ ] ] it is seen that the first term here is just the landauer formula , with the transmission coefficient averaged over the temporal period .
the second term forms a correction to this result , brought about by the modulating potential .
the pumped charge is given by @xmath91 ( see ref . ) , and it vanishes unless @xmath86 is time - dependent . clearly , a comparison between @xmath92 and the second part of @xmath93 is called for .
consider for simplicity zero temperature .
then , @xmath92 is given by the values of @xmath94 at @xmath95 , where @xmath96 is the chemical potential difference , and @xmath97 denotes the fermi level . on the other hand , the second term in @xmath93 is given by @xmath98 $ ] , and hence should be much smaller than the landauer contribution , which is proportional to @xmath99 . nevertheless , it may be of interest to explore this term experimentally , as it is related to the friedel phase of the nanostructure . to further explore this point , we consider the following simple example : a quantum dot , with a single localized level , coupled to two ideal 1d leads .
@xcite adopting the tight - binding description , we model the two leads connecting the quantum dot to the electronic reservoirs by 1d chains of sites , whose on - site energies are assumed to vanish , and whose nearest - neighbor transfer amplitudes are denoted by @xmath100 .
thus the energy of an electron of wave vector @xmath101 moving on such a chain is @xmath102 where @xmath103 is the lattice constant . the localized level , of energy @xmath104 , is attached to the left - hand - side lead with matrix element @xmath105 , and to the right - hand - side lead with matrix element @xmath106 .
the latter two quantities are assumed to vary slowly in time , in a periodic way .
our formalism requires just the knowledge of the instantaneous scattering matrix of the system . for the case at hand , @xmath107,\label{sng}\end{aligned}\ ] ] with @xmath108 let us consider first the pumped charge . at zero temperature , it is given by @xmath109,\end{aligned}\ ] ] in which energies are measure in units of @xmath110 , and @xmath111 are the time - dependent parameters of the system .
note that these two parameters can be thought of as the ` contact conductances ' of the quantum dot .
now imagine those to vary in time as follows : initially , both are equal to x@xmath112
. then x@xmath113 is increased linearly in time until it reaches the value x@xmath114 , while x@xmath115 is being held fixed at the value x@xmath112 . from that point , x@xmath113 is held fixed , while x@xmath115 increases linearly to the value x@xmath114 , and so on , see fig . [ fig1 ] .
= 8truecm it is quite straightforward to find @xmath92 for such a cycle .
one obtains @xmath116 with @xmath117 the resulting charge differs significantly from zero , and approaches unity ( in units of @xmath50 ) as long as the line of maximal transmission in the x@xmath113-x@xmath115-plane is well within the closed orbit forming the period .
@xcite in the present model ,
that line is given by @xmath118 we will elaborate on this point , which is exemplified in fig .
[ fig2 ] , in a future publication . for the parameters used to produce the results shown in fig .
[ fig2 ] , for x@xmath119 the maximal transmission line is contained within the pumping contour ( see fig .
[ fig1 ] ) , while for its higher values the pumping contour shifts away from the maximal transmission line .
= 8truecm let us next consider the time - average that appears in the second term of the biased current , i.e. , @xmath120 . in our example , this quantity becomes @xmath121 on a symmetric periodical curve like that presented in fig .
[ fig1 ] this integral vanishes .
hence , in such a symmetric configuration , there will be no deviation from the landauer formula due to the temporal variation of the friedel phase .
one may conclude from this example that when the pumping orbit in the plane of the time - dependent parameters is not so simple as the one considered above , there might be a small correction , related to the asymmetry of the orbit .
the motivation of this work is mainly to examine the validity of the adiabatic approximation in calculating the charge pumped quantum - mechanically through a confined system of non - interacting electrons , and to study the effects of a constant bias . to this end , we have derived the time - dependent current generated when the system is subject to a slowly - changing modulating potential , in addition to being connected to reservoirs of different chemical potentials .
we have found that the widely - used brouwer formula which gives the pumped charge in terms of the instantaneous scattering matrix is valid for arbitrary amplitudes of the modulating potential , as long as the lowest - order adiabatic approximation can be employed . in the process , we have obtained an alternative expression for the pumped charge , eq .
( [ intpump ] ) , which gives it in terms of matrix elements of the temporal derivative of the potential between the instantaneous scattering states .
we have also derived and analyzed the effects of the modulating potential on the biased current , and showed them to be quite small .
our formalism allows for the possibility to obtain systematically the corrections to the lowest - order adiabatic approximation .
we thank p. wlfle for helpful conversations .
this research was carried out in a center of excellence supported by the israel science foundation , and was supported in part by the national science foundation under grant no .
phy99 - 07949 ( at the itp , santa barbara ) , and by the albert einstein minerva center for theoretical physics at the weizmann institute of science .
the results presented above have been obtained in the adiabatic approximation , in which only first derivatives with respect to time are kept . as stated above
, this means that in the iterative solution of eq .
( [ chitild ] ) only the first iteration has been maintained .
here we discuss the contribution of the next iteration . for the sake of simplicity ,
we carry out this calculation for an un - biased system with two single - channel leads .
to second order in the temporal derivative , the scattering solution can be presented in the form @xmath122 with @xmath123 it follows that the quantity @xmath124 , eq . ( [ u ] ) ,
is now modified into @xmath125 as a result , we find that the current entering lead @xmath12 consists of two parts , the leading order in the adiabatic approximation , @xmath126 , which has been discussed above , and a correction , @xmath127 .
explicitly , @xmath128,\nonumber\\ \delta i_{\beta}(t)&=&e\int \frac{de}{2\pi}\sum_{\alpha}f_{\alpha}(e)\nonumber\\ & & \times\bigl [ -|u_{\beta\alpha}|^{2}+2\re\bigl ( s^{t\ast}_{\beta\alpha}\delta u_{\beta\alpha}\bigr ) \bigr].\label{deltai}\end{aligned}\ ] ] let us first verify that the correction @xmath127 obeys charge conservation over the entire period . to this end
, we sum @xmath127 over @xmath12 . using eq .
( [ sums ] ) , we have @xmath129\nonumber\\ & = & \int d{\bf r}\int d{\bf r}'\bigl [ \sum_{\beta}\chi^{t}_{\beta}({\bf r})\dot{\chi}^{t}_{\alpha}({\bf r})\chi^{t\ast}_{\beta}({\bf r}')\dot{\chi}^{t\ast}_{\alpha}({\bf r}')\nonumber\\ & -&2\im\bigl ( \chi^{t\ast}_{\alpha}({\bf r})\frac{d}{dt}g^{t}(e;{\bf r},{\bf r}')\dot{\chi}^{t}_{\alpha}({\bf r}')\bigr ) \bigr ] .\label{curcon}\end{aligned}\ ] ] we can now employ the fact that the current should be conserved upon integrating over the entire period , that is , when expression ( [ curcon ] ) is inserted into @xmath130 .
then we may integrate the second term by parts .
consequently , using the relation @xcite @xmath131 the two terms in ( [ curcon ] ) exactly cancel one another .
let us now turn to the expression for the correction @xmath132 , eq .
( [ deltai ] ) , and insert there the condition that the system is un - biased , i.e. , @xmath133 . then , using eqs .
( [ sums ] ) , ( [ chitt ] ) , ( [ dchide ] ) , and ( [ img ] ) , we find @xmath134 and @xmath135\nonumber\\ & + & \im\int d{\bf r}\bigl ( \frac{\partial^{2}\chi^{t}_{\beta}({\bf r})}{\partial e^{2}}\bigr ) \ddot{v}({\bf r},t)\chi^{t\ast}_{\beta}({\bf r}).\end{aligned}\ ] ] to obtain the last equality , we have made use of the relations @xmath136 and @xmath137 both relations are obtained by taking derivatives of eqs .
( [ green ] ) and ( [ chiteq ] ) with respect to the energy and the time .
collecting all these terms , we obtain @xmath138 it is satisfactory to note that , again , ( upon integrating by parts with respect to the energy ) , the energy integral includes the derivative of the fermi function .
hence , up to the second - order in the adiabatic approximation , we find @xmath139.\end{aligned}\ ] ] to estimate the relative magnitude of the correction ( the second term above ) compared to the leading order one ( the first term there ) , consider the part including the second derivative of the potential .
( [ dchide ] ) , this part becomes @xmath140 it follows that the correction will be smaller than the leading order term by a factor proportional to @xmath22 , due to the extra temporal derivative , times the energy derivative of the instantaneous scattering state .
the latter will include the energy - derivatives of the instantaneous transmission and reflection amplitudes ( which appear in @xmath141 ) , and possibly a term proportional to the 1d density of states , i.e. , the velocity , ( coming , e.g. , from the factors @xmath142 which appear in @xmath141 of our simple example discussed in sec .
iii ) . we may conclude that as long as the energy derivative of the instantaneous scattering solution is small on the scale of @xmath143 , the leading order adiabatic approximation suffices to obtain the pumped charge .
b. l. altshuler and l. i. glazman , science * 283 * , 1864 ( 1999 ) .
f. hekking and yu .
v. nazarov , phys .
b * 44 * , 9110 ( 1991 ) .
p. w. brouwer , phys .
b * 58 * , r10135 ( 1998 ) . i. l. aleiner and a. v. andreev , phys
81 * , 1286 ( 1998 ) .
f. zhou , b. spivak , and b. l. altshuler , phys .
lett . * 82 * , 608 ( 1999 ) .
t. a. shutenko , i. l. aleiner and b. l. altshuler , phys .
b * 61 * , 10366 ( 2000 )
. s. h. simon , phys .
b * 61 * , r16327 ( 2000 ) .
p. a. maksym , phys .
b * 61 * , 4727 ( 2000 ) . y. levinson , o. entin - wohlman , and p. wlfle , cond - mat/0010494 ( 2000 )
. y. wei , j. wang , and h. gou , phys .
b * 62 * , 9947 ( 2000 ) .
m. moskalets and m. bttiker , cond - mat/0108061 ( 2001 ) .
p. sharma and c. chamon , phys .
87 * , 096401 ( 2001 ) .
y. makhlin and a. mirlin , phys .
lett . * 87 * , 276803 ( 2001 ) .
y. wei , j. wang , h. guo , and c. roland , phys .
b * 64 * , 115321 ( 2001 ) .
m. switkes , c. m. marcus , k. campman and a. c. gossard , science * 283 * , 1905 ( 1999 ) .
h. pothier , p. lafarge , u. urbina , d. esteve and m. h. devoret , europhys . lett . * 17 * , 249 ( 1992 ) .
l. p. kouwenhoven , a. t. johnson , n. c. van der vaart , a. van der enden , c. j. p. m. harmans , and c. t. foxon , z. phys .
* b85 * , 381 ( 1991 ) .
d. j. thouless , phys . rev .
b * 27 * , 6083 ( 1983 ) .
m. bttiker , h. thomas , and a. prtre , z. phys .
* b94 * , 133 ( 1994 ) . j. e. avron , a. elgart , g. m. graf , and l. sadun , phys . rev .
b * 62 * , r10 618 ( 2000 ) .
a. aharony and o. entin - wohlman , cond - mat/0111053 ( 2001 ) .
y. levinson , phys .
b * 61 * , 4748 ( 2000 ) . o. entin - wohlman , y. levinson , and p. wlfle , phys . rev .
b * 64 * , 195308 ( 2001 ) . | a confined system of non - interacting electrons , subject to the combined effect of a time - dependent potential and different external chemical - potentials , is considered .
the current flowing through such a system is obtained for arbitrary strengths of the modulating potential , using the adiabatic approximation in an iterative manner .
a new formula is derived for the charge pumped through an un - biased system ( all external chemical potentials are kept at the same value ) ; it reproduces the brouwer formula for a two - terminal nanostructure .
the formalism presented yields the effect of the chemical potential bias on the pumped charge on one hand , and the modification of the landauer formula ( which gives the current in response to a constant chemical - potential difference ) brought about by the modulating potential on the other .
corrections to the adiabatic approximation are derived and discussed . 2 |
the differing properties of galaxies in high- and low - density regions indicate that there is a tug - of - war between the influence of internal processes and external factors on galaxy evolution , depending on the surroundings @xcite . in the higher density environment
, galaxies are subjected to external effects such as interactions or mergers which seem frequent and universal in the densely populated system , leading to enhanced star formation activity compared to field galaxies @xcite . in addition
, major mergers can deform the morphology of a galaxy @xcite causing a higher fraction of early - type galaxies in high - density environments @xcite .
on the other hand , the galaxies in less dense regions , like field galaxies , are thought to have experienced a relatively small number of mergers .
this low merger rate helps internal processes become significant because internal secular evolution , such as the formation of galactic bars and spiral structures , is too slow and fragile to be sustained against violent major mergers @xcite .
some studies , in this less harsh environment , found that early - type galaxies are observed to be less massive and relatively younger @xcite , and late - type galaxies prefer to have some internally developed features such as a pseudo - bulge @xcite , confirming the differing evolution of galaxies in low - density regions .
an isolated galaxy is considered as an object which has not been considerably perturbed by its neighbors for a few gyr @xcite .
therefore , in the context of the local environment effect , many studies of isolated galaxies have been conducted in the last few decades @xcite .
one of the well known studies is the analysis of the interstellar medium of isolated galaxies ( amiga ) project whose galaxies are based on the catalogue of isolated galaxies ( cig ; * ? ? ?
the amiga project refined the cig for morphologies of 1,050 isolated galaxies @xcite using the second palomar observatory sky survey ( poss ii ) .
the degree of isolation was also quantified @xcite using the digitized poss ( dposs i and ii ) . in a series of studies , the amiga project dealt with multiwavelength properties of isolated galaxies including photometric characteristics @xcite , the star formation rate derived from h@xmath2 emission @xcite , the fir luminosity function @xcite , radio continuum properties @xcite and the nuclear activity using fir , radio and optical databases @xcite . in defining an isolated galaxy ,
the isolation criteria devised by @xcite , or variations on this theme , are widely used in studies including amiga projects .
@xcite only took into account a relative apparent size and a projected distance between the target galaxy and potential neighbors , due to the deficit of redshift information at that time .
but the photometric isolation scheme can miss some isolated galaxies by regarding background objects as potential companions @xcite .
in addition to this , tidal perturbation strength can be considered as a criterion of isolation @xcite .
the tidal perturbation strength helps to quantify the tidal influence of the companions based on spatial separations and stellar mass ratio between the target galaxy and companions .
for example , if external tidal forces are greater than 1% of the internal binding force , the galaxy can be thought to be perturbed by neighbors @xcite .
however , the external influence may not be instantaneous but continuous , so accumulated effects during the transit time of neighboring galaxies should be considered to better understand the effects of perturbers .
statistical studies of isolated galaxies using the sloan digital sky survey ( sdss ; * ? ? ?
* ) have been reported @xcite .
many of them are based on the criteria developed by @xcite , and only focused on providing a catalog of isolated galaxies or testing the cig criteria . in this paper , we constructed a catalog of isolated galaxies using the sdss data release 7 ( dr7 ; * ? ? ? * ) to present the properties of isolated galaxies along the hubble morphological type . for the isolation criteria
, we took into account the distance in three dimensional space to reduce the impact of projection effects .
we did not use the tidal perturbation strength , but instead rely on a strict distance criterion . in section [ sec : sample ] , we present our sample selection criteria for isolated galaxies and comparison sample galaxies . in section [ sec :
basic ] , we describe the morphologcal classification scheme and photometric properties of isolated galaxies . the characteristics of absorption lines with a stellar population model and the properties of emission lines are addressed in sections [ sec : absorption ] and [ sec : emission ] , respectively . in section
[ sec : discussion ] , we discuss the evolutionary scenario of isolated galaxies compared to comparison galaxies and possible biases in the sampling .
we present our conclusions and summarize our results in section [ sec : summary ] . throughout this work ,
we have assumed the following cosmological parameters : h@xmath3 = 70 mpc@xmath4 , @xmath5 = 0.3 and @xmath6 = 0.7 .
to study isolated galaxies , we make two kinds of samples : isolated galaxies without any companions , and a comparison sample of galaxies in denser regions .
the parent sample consists of the main galaxy sample @xcite of the sdss dr7 which covers one quarter of the sky using a 2.5 m telescope in apache point , new mexico @xcite .
we focus on the nearby galaxies ( @xmath7 ) so as to include more faint galaxies when we measure the galaxy density in our volume - limited sample .
this redshift range also helps us to secure relatively robust visual morphological classification while minimizing the evolution effect , as described by @xcite .
the main galaxy sample is comprised of galaxies with spectroscopic information and most of them are at @xmath8 .
the sample is designed to be complete for galaxies with @xmath9 in spectroscopy .
however , galaxies with low surface brightness or galaxies which are too bright in apparent fiber magnitudes are not targeted for the spectroscopic survey @xcite .
furthermore , some galaxies in crowded fields can be missed due to the fiber collisions @xcite .
therefore , before defining isolated galaxies , we test the spectroscopic completeness for each field of our galaxies . the spectroscopic completeness of each field is estimated by calculating the percentage of galaxies with spectra among galaxies in the luminosity range of @xmath10 , a projected 2 mpc radius around our target galaxy . 2 mpc is our criterion of isolation ( see below ) .
since the spectroscopic completeness declines for bright galaxies , we set the minimum @xmath11-band magnitude to be 13.0 which guarantees at least 70% of the spectroscopic completeness @xcite .
we have found that only about 10% of galaxies in @xmath7 ( with the mean completeness of 89% ) have the completeness less than 70% and they are removed at this stage .
the isolated galaxies and the comparison group galaxies used in our study are selected based on a volume - limited sample ( @xmath12 ) .
this magnitude cut roughly corresponds to a stellar mass of @xmath13 .
after the volume limitation , we have 38,011 galaxies . for this volume - limited sample
, we measure the density of each galaxy by counting the number of companion galaxies within a sphere in three dimensional space . in estimating a distance between galaxies using the redshift
, the distance could be overestimated due to proper motions of field galaxies ( @xmath14190 ; * ? ? ?
* ) , misclassifying some galaxies to be isolated . in other studies , an elongated cylinder with 0.5 or 1 mpc radius
is considered to compensate for the effect of the proper motion in selecting isolated galaxies @xcite . since we define isolated galaxies based on a sphere instead of a cylinder , a 2 mpc distance criterion in redshift space , which is a strict condition ,
is chosen .
this 2 mpc criterion can help companion galaxies to be separated from the target galaxy by at least 1 mpc in `` real '' space .
for example , in the worst case scenario , the target and companion galaxy are aligned along the line of sight .
we take the case that the galaxies appear separated by 2 mpc in redshift space , but in fact the companion galaxy has a proper motion of 190 .
we calculate that if we vary the direction of the proper motion of the companion randomly , the companion will have a true separation from the target galaxy greater than 1 mpc , 70% of the time ( prob = @xmath15 = @xmath16 = 0.7 ) .
as this is a worst case scenario , we can expect our 2 mpc criterion to guarantee at least a 1 mpc true separation in most cases . to choose the comparison galaxies as the `` typical '' galaxies according to the hierarchical formation scenario , we select the comparison galaxies when there are at least two companion galaxies in a 1 mpc sphere under the consideration of the finger - of - god effect @xcite .
a typical system in the comparison sample is similar to the local group the milky way has two companion galaxies within 1 mpc , m31 and m33 .
when we account for companion galaxies , the mass ratio between the target galaxy and companion galaxies varies depending on the mass of the target galaxy . for example , massive galaxies can have a large range of mass ratio .
meanwhile , for galaxies near the magnitude limit ( @xmath17 ) , only galaxies with comparable mass can be considered as companion galaxies . in other words ,
our definition of isolation could permit some `` isolated galaxies '' to have companion galaxies with comparable mass below the magnitude cut off .
to minimize this inhomogeneity in the sampling , we select galaxies with @xmath18 ( 1 magnitude brighter than the magnitude limit ) .
this ensures companions can be detected near the magnitude limit down to mass ratio of @xmath141:3 ( the mass limit of target galaxies is @xmath14@xmath19 @xmath20 and that of companion galaxies is @xmath14@xmath21 @xmath20 ) . with this choice ,
the number of isolated galaxies and comparison sample galaxies are 3,969 and 4,591 , respectively .
to examine the surroundings of our galaxies , it is helpful to visualize the spatial distribution of the galaxies . figure [ fig : spatial ] shows a column density map with isolated and comparison sample galaxies .
we measure the gaussian weighted density of galaxies for our volume - limited sample ( 38,011 ) @xcite : @xmath22 @xmath23 is the angular distance and @xmath24 is the distance along the line of sight .
@xmath25 is introduced to compensate for the finger - of - god effect .
we choose @xmath26 to be 3 mpc to present long - range densities .
the darker shading indicates the higher long - range density and isolated and comparison sample galaxies are shown in blue and red symbols , respectively .
a significant fracton of the volume is empty space ( voids ) .
relatively denser regions form filamentary structures .
most of comparison sample galaxies seem to reside in the vicinity of the high - density region .
meanwhile , isolated galaxies are widely distributed from relatively less dense regions to higher density areas @xcite .
it is clear that the definition of isolated galaxies in our sample is with respect to the local density , rather than the global environment .
it is interesting to compare our galaxies with void galaxies which are regarded as galaxies in very low - density regions . for a void catalog
, we adopted the catalog provided by @xcite .
@xcite tried to identify voids in sdss dr 7 using a void finder @xmath27 which is based on the voronoi tessellation .
the @xmath27 algorithm is liable to include the surrounding high - density walls of the void @xcite . therefore , to avoid including the galaxies in void walls , we only select the members located in the central regions of voids by using smaller effective size of voids than the given value , i.e. 0.8r@xmath28 . as a result , we find 59 voids in our sampling volume with a median radius of 7 mpc , and @xmath143% of isolated galaxies and @xmath144% of comparison galaxies are in voids
. comparable numbers of void galaxies in the isolated and comparison sample indicates that the void galaxies can be either isolated or coupled .
also , a small number of void galaxies in our sample implies that void galaxies are in `` globally '' low - density regions while our isolated galaxies are in `` locally '' low - density areas .
thus the isolated galaxies can be expected to show properties caused by local environments , rather than global ones .
it is hard to classify the morphology of galaxies that appear small or edge - on .
thus , we take advantage of the criteria used in @xcite to remove apparently small or low isob@xmath29/isoa@xmath29 objects among isolated and comparison sample galaxies : isoa@xmath29 @xmath30 30@xmath31 or isob@xmath29/isoa@xmath29 @xmath30 @xmath320.01 isoa@xmath29 + 0.90 . here , isoa@xmath29 and isob@xmath29 from the sdss photometric pipeline are the isophotal radius in @xmath11-band along the major - axis and minor - axis , respectively . to test the criteria
, we also perform a morphological classification using a randomly selected 500 galaxies from our sample , and confirm that these demarcation lines separate the classifiable and non - classifiable objects well .
this step removes about 30% of galaxies , leaving 2,660 isolated galaxies and 3,409 comparison sample galaxies for a visual inspection .
when we exclude the galaxies which can not satisfy the criteria , some galaxies with intrinsically low isob@xmath29/isoa@xmath29 like e6 and e7 are ruled out , making a bias against them .
therefore , we will not address such galaxies in our study .
3 @xmath33 composite color images with different scales , and one gray scale image from the sdss are used in the visual inspection .
the field - of - view of the 3 colored images are 50 , 70 and 100 kpc .
we classify the morphology as elliptical , lenticular , unbarred spiral , barred spiral and irregular galaxy following the hubble sequence @xcite .
galaxies which can not be placed in a specific class are categorized as `` unknown '' type galaxies .
we did not consider other revised versions of hubble types such as classifications used in third reference catalog of bright galaxies ( rc3 ; * ? ? ?
* ) to keep simplicity .
the morphology classification strategy we use is similar to the classification method described in @xcite .
here is a brief description of the scheme . 1 .
we classify a galaxy as an elliptical galaxy when it is a spheroidal dominant system without any distinct disk or clumpy features .
elliptical galaxies are divided into subclasses depending on the projected ellipticity defined as @xmath34 = 1@xmath32isob@xmath29/isoa@xmath29 .
these subclasses are only derived from the projected shape and thus may not be highly indicative of the true shapes @xcite .
lenticular galaxies are bulge dominant but also have a disk feature . with the exception of the presence of a disk , the properties of lenticular galaxies are similar to those of elliptical galaxies , making it hard to distinguish between them @xcite
. however lenticular galaxies can be distinguished by a less concentrated and extended light distributions compared to elliptical galaxies .
the most distinct feature of spiral galaxies is the presence of spiral arms . among the spiral galaxies , early - type spiral galaxies such as sa or sba
have tightly wound spiral arms and a relatively large bulge . from s(b)a to s(b)d , spiral arms are wound less tightly and the bulge becomes less dominant .
furthermore , the presence or absence of a bar is also an important characteristics of spiral galaxies . by using the gray scale images with sharper contrast
, we try to minimize the bias caused by weak bars , which are difficult to identify due to their low luminosity and short length @xcite .
4 . irregular galaxies have a wide variety of forms . in general , they show weak spiral arm features and a very small bulge ( or maybe bulgeless ) , and are rather asymmetric .
galaxies in the unknown class are the most difficult to define
. the majority of them may be lenticular galaxies which tend to be confused with elliptical galaxies , as mentioned above .
many small ( though brighter than our limit ) galaxies , without any distinct feature , were difficult to classify and thus placed into the unknown category .
very faint galaxies , with few striking features , and also galaxies that are disturbed by tidal interactions or mergers , are also included in the unknown category .
[ tab : catalogfour].,scaledwidth=50.0% ] by excluding unknown galaxies from each sample , our sample has 1,644 isolated galaxies and 1,994 comparison sample members .
the similarity in size of the two samples makes our statistical analysis straightforward .
figure [ fig : fraction ] shows the distribution of morphology fraction along the hubble sequence .
the fraction of isolated galaxies and comparison sample galaxies are presented in gray and white bars , respectively .
e@xmath35 indicates the aggregates of subclasses of elliptical galaxies ( e0e5 ) .
early - type galaxies are more common in the comparison samples .
this is in agreement with the well - known morphology - density relation that high density environments are rich in elliptical galaxies , while low density regions are spiral - rich @xcite .
specifically , elliptical galaxies are more abundant in the comparison samples than in isolated galaxies by a factor of 3 ( 28.6% of comparison sample and 9.9% of isolated galaxies ) .
since the work of @xcite , it is generally accepted that a merger of two disk galaxies can result in the formation of an elliptical galaxy @xcite .
therefore , assuming the merger rate is proportional to the density of the environment , it is natural that the fraction of elliptical galaxies is higher in the comparison sample .
we must note , however , that the comparison sample galaxies are in diverse environments spanning from a small galaxy group to a large galaxy cluster . in a group environment , fly - by tidal interactions and mergers may occur frequently @xcite . on the other hand , in the cluster environment ,
mergers are unlikely to occur due to the fast relative motion of galaxies within the deep gravitational potential well of the cluster @xcite .
interestingly , deep optical observations show the frequency of the merger features in clusters is almost comparable to that in the field @xcite .
this contradiction seems to be explained by the merger relics scenario suggested by @xcite . in this scenario , galaxies with merger features flow into the cluster , having experienced their merger events outside the cluster environment . while complexities exist ,
it is probably natural to assume that today s dense regions were assembled by smaller density peaks in the earlier universe where mergers were frequent . in both samples , slightly oval shaped galaxies such as e2 or e3 are greatest in abundance and this abundance decreases as galaxies become rounder or flatter , confirming the result of @xcite . according to @xcite , luminous elliptical galaxies are likely to be triaxial and less luminous elliptical galaxies are usually oblate .
however , we can not find any prominent distinction between the two samples .
it is probably because our classification scheme does nt reflect their intrinsic shapes , but merely the projected ones .
.,scaledwidth=50.0% ] sa galaxies are more frequently found in the comparison sample , but sc galaxies are more abundant in isolated galaxies ( 7.9% , 19.7% , 15.2% in comparison samples and 5.8% , 26.3% , 30.1% in isolated galaxies for sa , sb and sc , respectively ) .
this trend is the same in barred spiral galaxies .
furthermore , for the isolated galaxy sample , there is a steady increase in number as we move from sa to sc galaxies , which is consistent with other studies @xcite . on the other hand
, comparison sample galaxies do not show such an increasing trend from sa to sc type .
this implies that the evolutionary histories of spiral galaxies in the isolated systems and relatively denser regions may be different .
the mechanism for bar formation is debatable .
the bar structures can be formed through either a secular evolution @xcite or during a tidal interaction with another galaxy @xcite .
thus , it is interesting to compare the bar fractions @xmath36 of the two samples in different density regions .
@xcite reported that the bar fraction in the clusters is comparable to that of field galaxies at lower redshift , whereas @xcite found no environmental preference .
furthermore , @xcite found the fraction of barred galaxies are not influenced by the nearest neighbor galaxy if the separation between galaxies is larger than 0.1 times the virial radius of the neighbor . on the other hand , @xcite claims that bars are more frequently found in perturbed late - type galaxies than in isolated ones by a factor of 2 .
we found that 18.7% and 18.1% of spiral galaxies are barred spiral galaxies in the isolated system and comparison sample , respectively / isoa@xmath37 0.7 as used in @xcite for example , the bar fraction increases up to 23.8% and 25.1% for isolated and comparison sample galaxies , respectively .
these values are close to the bar fraction found in other studies which use a visual inspection for the bar identification ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we acknowledge this systematic uncertainty , and we thus focus only on the difference between isolated and comparison samples . ] .
this result indicates that there is no significant difference in the bar fraction as a function of density in our sample .
while we admit that our environmental criterion is not comprehensive , one may be tempted to interpret this result as evidence against the environmental scenario for the bar formation .
the concentration index is widely regarded as a tool for morphology classification @xcite .
since we classified the morphologies of galaxies using a visual inspection , the concentration index can be used as a validity check for our classification .
there are several distinct definitions of concentration index , but here we use the ratio between petror90 and petror50 of the @xmath11-band petrosian magnitude , i.e. c@xmath29 @xmath38 petror90/petror50 . figure [ fig : cr ] shows c@xmath29 values on the hubble tuning fork .
the bar indicates the median value and the error bar is the 1@xmath26 dispersion .
elliptical galaxies usually have a concentration index greater than 3 with small dispersion , while late - type galaxies typically have smaller c@xmath29 . for all hubble - types
, there is no significant difference in concentration index between the isolated and comparison galaxies .
a similar test can be performed on @xmath39 in @xmath11-band which is a photometric parameter provided by sdss .
this value indicates the fraction of flux which the de vaucouleurs profile accounts for .
therefore , an elliptical galaxy tends to have @xmath39 close to unity , whereas a spiral galaxy has a lower @xmath39 . in figure
[ fig : fracdev ] , the median values of @xmath39 of elliptical galaxies are close to 1 , regardless of the subclasses . for late - type galaxies , @xmath39 decreases from s(b)a to s(b)d , as expected . and also
, like c@xmath29 , we can not find any prominent difference in @xmath39 between our different density samples . in this section
, we compared our morphological classification result with those of other studies .
@xcite classified the morphology of 10,233 galaxies for the sdss dr7 database using a visual inspection based on the hubble classification scheme .
their redshift range ( @xmath40 ) is similar to ours , and thus there is some overlap with our sample . among isolated galaxies ( 2,660 ) and comparison sample galaxies ( 3,409 ) , 1,575 and 1,965 galaxies
are matched with the galaxies in @xcite , respectively .
similarly , @xcite performed visual inspections for 14,034 galaxies in the sdss dr4 following their own classification scheme which is similar to rc3 classifications under the consideration of fine structures like rings , lenses and shells .
they use galaxies with an apparent limit of g @xmath41 mag , and in the redshift range of @xmath42 which covers our sample s redshift range . as a result
, 936 isolated galaxies and 1,113 comparison sample galaxies of our galaxies overlap with their sample .
figure [ fig : compare ] shows the result of morphological comparison for isolated galaxies ( upper panels ) and comparison sample galaxies ( lower panels ) .
the percentage represented by gray scale shading is calculated by counting the number of galaxies that match the morphological type in the other studies . on the whole ,
our classification is in good agreement with others .
for all morphologies except the unknown type , @xmath1485% of galaxies in each panel are near the one - to - one line .
there are a number of galaxies classified as s0 galaxies in @xcite but as unknown in our scheme , meaning that we were more conservative in our classification of s0 galaxies ( see section [ sec : morph ] ) than @xcite . , 1.0@xmath26 , and 2.0@xmath26 are drawn.,scaledwidth=50.0% ]
depending on the morphological type , the galaxies tend to reside in different regions on the color - magnitude diagram .
the magnitude and color are galactic extinction - corrected with the data from the sdss pipeline and @xmath43-corrected using @xmath44 v4.2 package @xcite .
early - type galaxies lie on a well - defined red - sequence , whose tilt shows that massive early - type galaxies are redder than less massive ones @xcite . in figure
[ fig : cmr ] , for both isolated galaxies ( solid line ) and comparison sample galaxies ( shade region ) , there are tight color - magnitude relations ( cmr ) of elliptical galaxies .
elliptical galaxies in the isolated system are fainter ( @xmath45 ) , on the whole , than those in high density environments ( @xmath46 ) with an offset in median @xmath47 by @xmath140.4 .
this result can be understood within the hierarchical merger scenario if galaxies in denser regions have more opportunity to become massive through mass accretion and mergers ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
furthermore , for the galaxies with @xmath48 , @xcite found that the nuv - r color of early - type galaxies in low - density environments extends further into the blue than that of early - type galaxies in high - density environments , indicating the presence of residual star formation @xcite .
however , we can not find a prominent vertical offset in color , probably because the ( g r)@xmath3 color is not very sensitive to the residual star formation @xcite .
unbarred spiral galaxies prefer to be located in the blue cloud as opposed to in the red sequence @xcite .
unbarred spiral galaxies for both isolated and comparison sample are in the blue cloud , and the overall distributions are similar ( the lower panel of figure 6 ) .
the offset in median @xmath47 ( @xmath140.15 ) between isolated and comparison sample galaxies is less significant than that found in elliptical galaxies .
also , the difference in median ( g - r)@xmath3 color between the two samples is only @xmath140.03 which may result from the different median @xmath47 .
this value is consistent with that found in the literature @xcite .
in contrast to unbarred spirals , barred spiral galaxies show a bimodality with two peaks ; a luminous and red peak , and a faint and blue peak ( middle panel of figure [ fig : cmr ] ) .
this is consistent with other studies ( @xcite ; @xcite ) , although the strength of the second peak can vary depending on the methods for identifying a weak bar or the sample selection criteria @xcite .
the bimodality appears in both the isolated and comparison sample galaxies .
there is no obvious difference between the contours of the two samples , nor in the bar fraction ( see section [ sec : mor_distribution ] ) .
.,scaledwidth=50.0% ] the stellar mass is calculated using the formula in @xcite with the @xmath43-corrected g
r color and @xmath11-band magnitude .
the stellar mass along the hubble tuning fork is shown in figure [ fig : mass ] .
the median value of the stellar mass of early - type galaxies is higher than that of late - type galaxies in both isolated and comparison sample as expected . among elliptical galaxies , there is no big difference among the subclasses , because the subclass is not revealing the true geometry due to projection effects @xcite . in late - type galaxies ,
the stellar mass becomes smaller as the galaxies tend to be less bulge dominated for both barred and unbarred spiral galaxies @xcite , regardless of the surrounding environmental density .
the median stellar mass of elliptical galaxies ( e@xmath35 ) in the comparison sample ( @xmath49 ) is about 50% larger than that of isolated elliptical galaxies ( @xmath50 ) .
this can be understood as a result of difference in the merger and accretion history as mentioned previously . meanwhile , for the stellar mass of late - type galaxies , there are less prominent distinctions ( 1020% ) between the isolated galaxies and comparison sample galaxies along hubble types .
this may imply that the majority of mass of late - type galaxies come from the in - situ star formation accompanied with a minor contribution from the merger @xcite , regardless of the environment density .
the results are consistent when we instead use the stellar mass from the mpa - jhu catalog .
.,scaledwidth=50.0% ] we use the velocity dispersion provided by the ossy catalog @xcite , that were measured using @xmath51 based on the improved spectral fits .
since the sdss performed the spectroscopy using fibers with 3@xmath31 diameter , we use the effective radius of galaxies @xcite in order to obtain the effective velocity dispersion @xcite , corrected for the size of the fiber .
we chose only galaxies with @xmath52 and @xmath53/@xmath54 .
as a result , we have 1,386 ( 84% ) and 1,816 ( 91% ) galaxies for the isolated and comparison sample , respectively .
most of the early - type galaxies ( 99.9% ) and s(b)a and s(b)b ( 92% ) satisfy this condition , but plenty of s(b)c and s(b)d ( 68% remain ) are removed due to their intrinsically low velocity dispersions making it difficult to measure with high precision .
whether in the isolated or comparison sample , elliptical galaxies have larger velocity dispersions than late - type galaxies ( figure [ fig : sigma ] ) , which is consistent with previous studies @xcite . for elliptical galaxies , the median value of the velocity dispersions of comparison sample galaxies ( 194@xmath55 ) is higher than that of isolated galaxies ( 160@xmath56 ) .
however , the velocity dispersions of spiral galaxies shows only small differences between isolated and comparison sample galaxies .
this trend is similar to that of the stellar mass because the velocity dispersion is related to the dynamical mass , which may roughly trace the stellar mass .
absorption lines of galaxies are essential tools to study their stellar populatons , which in turn provide information on the history of a galaxy s formation and evolution . in this study
, we use measurements of absorption lines given by the ossy catalog @xcite to study stellar populations .
the ossy catalog , based on the lick / ids system resolution , provides the strength of absorption lines which are measured taking into consideration for the emission line subtraction using the @xmath51 fit tool .
accurate subtraction of emission lines is important , because emission lines near absorption lines are likely to hinder determining the pseudocontinuum properly .
we apply a signal - to - noise cut ( @xmath57 ) and quality - assessing parameter ( n@xmath58 ) to make the absorption line strength more reliable . as a result , 98.4% and 98.8% of galaxies for the isolated and the comparison sample remained , respectively . as
the line measurements only represent the spectroscopic properties of the central part of galaxies due to the size of sdss fiber ( 3 ) , they can not be compared directly with the photometric characteristics .
however using both sets of data can help to understand the overall properties of the galaxies .
absorption line strengths .
the format is the same as that of figure [ fig : cr].,scaledwidth=50.0% ] the strength of absorption lines are generally affected by temperature and metallicity . among the absorption lines , balmer absorption line indices are sensitive to the temperature , and strongest around an effective temperature of 10,000 k ( a - type stars ) .
therefore , an old stellar population which has a relatively low temperature shows a weak balmer absorption line in general @xcite . for the analysis of balmer absorption line , we use the h@xmath1 absorption line .
equivalent width ( ew ) values of the h@xmath1 absorption line are shown in figure [ fig : hb ] .
the h@xmath1 absorption line strength of early - type galaxies is lower than that of late - type galaxies @xcite , indicating that the stellar component of early - type galaxies is older , as expected .
also , in late - type galaxies , the median ew of h@xmath1 absorption line becomes larger along the hubble sequence for both isolated and comparison sample galaxies .
this is consistent with the idea that more late - type spiral galaxies are more gas rich , with a higher star formation rate ( sfr ) , and thus relatively younger .
although the median ew of the h@xmath1 absorption line index in isolated elliptical galaxies ( 1.83@xmath59 ) is slightly larger than that of elliptical galaxies in the comparison sample ( 1.75@xmath60 ) and this trend happens for all subclasses of elliptical galaxies , the differences are not notable . for spiral galaxies , they do nt show any prominent difference .
since the strength of the h@xmath1 absorption line is also affected by metallicity , a more detailed analysis will be performed in section [ sec : sec_pop ] .
.,scaledwidth=50.0% ] according to @xcite , massive galaxies with deep gravitational potential wells can better maintain their interstellar media against ejection by sne explosions , leading to more efficient chemical recycling . therefore in this scenario ,
massive galaxies are expected to be more metal rich , as observed in the well - known mass metallicity relation ( e.g. , * ? ? ?
to study the metallicity , we chose the fe5270 absorption line index as a representative metal line which is a typical fe indicator used in the stellar population model @xcite . figure [ fig : fe5270 ] shows , for both isolated and comparison samples , early - type galaxies have larger ew values of the fe5270 absorption line than late - type galaxies do .
this can be inferred from the fact that early - type galaxies are more massive than late - type galaxies ( figure [ fig : mass ] ) resulting in higher metallicity . furthermore , late - type galaxies have descending trends of the fe5270 line strength from s(b)a to s(b)d , regardless of the density environment , and this can also be understood in terms of the mass - metallicity relation . like h@xmath1 , there is a small difference of fe5270 line strength between isolated and comparison sample galaxies , for both elliptical and spiral galaxies , but it is not significant enough to compare their metallicity without consideration of the degeneracy of age and metallicity .
.,scaledwidth=50.0% ] type ii supernovae are thought to occur shortly after the onset of star formation . as a result
, they quickly spread @xmath2-elements , their main product , into the interstellar medium .
meanwhile , type ia supernovae whose generation is delayed , but relatively continuous throughout the period of star formation , supply predominantly fe - peak elements @xcite .
thus , by comparing the abundance of @xmath2-elements and fe , [ @xmath2/fe ] , we can trace the star formation history of galaxies @xcite .
the mgb absorption line is widely used to represent the abundance of @xmath2-elements .
we can see the mgb absorption line strength of early - type galaxies is larger than that of the late - type galaxies .
also among late - type galaxies , the strength decreases from s(b)a to s(b)d for both isolated and comparison sample galaxies ( figure [ fig : mgb ] ) .
this result implies that the early - type galaxies experience a relatively intensive star formation activity , while the late - type galaxies have an extended star formation history . for elliptical galaxies ,
the median ew of mgb of isolated galaxies ( 4.02@xmath61 ) is a bit smaller than that of comparison galaxies ( 4.32@xmath62 ) , but the difference is not large . for spiral galaxies ,
the difference is even less significant .
as we discussed above , absorption line strengths are affected by both age and metallicity , the so called ` age metallicity degeneracy ' @xcite .
thus , it is important to consider the absorption line indices such as h@xmath1 , fe5270 and mgb simultaneously , based on the stellar population model , in order to try and break the degeneracy . in this section
, we are going to discuss the stellar age , metallicity and [ @xmath2/fe ] derived from the stellar population model devised by @xcite ( hereafter the tmj model ) .
the stellar population models are constructed based on an evolutionary population synthesis model with calibrations ( @xcite ; @xcite ) .
we used the tmj model which covers a wider range of [ @xmath2/fe ] values than the model of @xcite ( tmb model ) used in @xcite .
we should bear in mind that this method provides the `` luminosity - weighted '' age and metallicity , and thus it is biased towards the younger , brighter stellar components . to break the degeneracy between age and metallicity ,
the model grid of the [ mgfe]@xmath63h@xmath1 plane was used .
we adopted the index [ mgfe]@xmath63 which is less sensitive to [ @xmath2/fe ] and therefore a good tracer of metallicity @xcite .
we fixed [ @xmath2/fe ] as 0.3 for simplicity , but this does not affect the results significantly because neither h@xmath1 @xcite nor [ mgfe]@xmath63 is responsive to [ @xmath2/fe ] .
the median values of the luminosity - weighted age and metallicity for each morphology are shown in figure [ fig : mgfe - hb ] .
we did not separate elliptical galaxies into subclasses .
the blue symbols indicate isolated galaxies and the red ones are comparison sample galaxies . along the hubble sequence ,
the median age and metallicity decrease for both isolated and comparison sample galaxies .
there is a trend that isolated galaxies are slightly younger and more metal poor based on median values than comparison sample galaxies for almost all morphologies , but the differences are not significant ( figure [ fig : mgfe - hb ] ) .
we can also investigate the properties of stellar populations using the mgb@xmath30fe@xmath64 plane for fixed ages .
here , @xmath30fe@xmath64 is defined as a mean value of fe5270 and fe5335 absorption lines , ( fe5270+fe5335)/2 @xcite . for a representative age , we chose the age 8 gyr for early - type galaxies including s(b)a and 3 gyr for late - type galaxies ( figure [ fig : mgb - fe ] ) .
as with the trends of the median age and metallicity derived from the [ mgfe]@xmath63h@xmath1 plane , the median [ @xmath2/fe ] decreases along the hubble sequence
. however , the median values of [ @xmath2/fe ] of isolated galaxies do not seem to be significantly different from those of comparison sample galaxies .
metallicities derived from the mgb@xmath30fe@xmath64 plane also show a weak offset between the two samples as seen in figure [ fig : mgfe - hb ] . for a quantitative analysis , we derived the luminosity - weighted age , [ z / h ] and [ @xmath2/fe ] for each galaxy using an iterative method .
this method is based on a circular logic ; we firstly estimate the age for fixed [ @xmath2/fe ] on the [ mgfe]@xmath63h@xmath1 plane , and then using the estimated age , derive the [ @xmath2/fe ] on the mgb@xmath30fe@xmath64 plane . by iterating this process for each galaxy
, the quantity converges to a single value .
figure [ fig : age ] shows contours ( left panels ) for luminosity - weighted ages of isolated ( orange solid line ) and comparison galaxies ( shade region ) along the effective velocity dispersion . for a fixed velocity dispersion
, the luminosity - weighted age does not differ significantly between the two samples .
however , the contour of isolated elliptical galaxies is notably extended to lower velocity dispersions and younger ages than that of their comparison galaxies .
therefore , the median luminosity - weighted age ( a representative age of the group ) is likely to be younger in the isolated elliptical galaxies than the comparisons .
this result is consistent with other studies which argue that , on average , the elliptical galaxies in low - density systems tend to be less massive and younger than those in high - density regions @xcite .
other morphologies seem to be almost identical at all contour levels .
right panels in figure [ fig : age ] show the fraction of galaxies corresponding to each age bin .
the kolmogorov - smirnov
( k - s ) test is also performed to address the similarity of distributions of isolated and comparison sample galaxies . for elliptical galaxies ,
the distributions of the luminosity - weighted age of the two samples seem slightly different .
if we take the significance level as 0.05 ( or 5% ) which is typically used , the small p - value of elliptical galaxies ( 0.003 ) suggests the younger median luminosity - weighted age of isolated galaxies ( 8.3 gyr ) than that of comparisons ( 10.0 gyr ) is statistically significant .
in addition , sc - type ( sc and sbc ) spiral galaxies have the p - value ( 0.04 ) smaller than 0.05 , implying there is a chance that the distributions of isolated and comparison galaxies arise from different parent samples .
however , there is no obvious difference in the median luminosity - weighted age of isolated galaxies ( 2.0 gyr ) and comparison ones ( 1.9 gyr ) . for sa - type ( sa and sba ) and sb - type ( sb and sbb ) spiral galaxies , the p - values are 0.235 and 0.688 , respectively and we can not see any notable differences in ages between isolated and comparison galaxies in each histogram .
the results of metallicity with the effective velocity dispersion are presented in figure [ fig : z_h ] .
it is difficult to discern any discrepancy between isolated galaxies and comparison samples for all morphologies in the contour diagram , except for slightly lower velocity dispersions of ellipticals in isolated galaxies .
also , histograms of the metallicity distribution for isolated and comparison galaxies seem similar for all panels .
the k - s test for the metallicity distribution indicates sa - type spiral galaxies have small p - values .
but the median luminosity - weighted metallicities of isolated and comparison galaxies do not show any prominent differences .
the situation of [ @xmath2/fe ] is similar .
only elliptical galaxies show contours slightly extended to lower velocity dispersions in isolated galaxies and there is no significant difference between isolated and comparison galaxies for late - type galaxies in contour diagrams ( figure [ fig : a_fe ] ) .
the p - value derived from the k - s test implies late - type galaxies are unlikely to have different [ @xmath2/fe ] distributions for those two samples .
meanwhile , the lower p - value of elliptical galaxies suggest different distributions , although there is no significant discrepancy in the median values between the isolated and comparison galaxies .
unlike absorption lines which indicate the cumulative star formation history , emission lines are more suitable for revealing ongoing star formation . furthermore , active galactic nuclei ( agns ) which show strong emission lines are thought to play an important role in galaxy evolution through agn feedback .
therefore , we now consider the properties of the emission lines .
we use the flux and ew values of emission lines from the ossy catalog .
lccccccccccccc + * [ -0.5ex]emission - line galaxies & 911 & 0.75 & 1.81 & 3.56 & 18.06 & 20.12 & 0.75 & 2.81 & 4.25 & 3.25 & 0.12 & 1.19 & 56.69 + & ( 796 ) & ( 3.20 ) & ( 1.67 ) & ( 3.25 ) & ( 13.70 ) & ( 11.42 ) & ( 0.36 ) & ( 2.44 ) & ( 2.18 ) & ( 1.27 ) & ( 0.20 ) & ( 0.61 ) & ( 40.28 ) + + * [ -0.5ex] star - forming & 633 & 0.00 & 0.38 & 0.94 & 12.94 & 17.44 & 0.75 & 0.56 & 2.25 & 2.75 & 0.12 & 1.19 & 39.31 + & ( 486 ) & ( 0.05 ) & ( 0.15 ) & ( 0.86 ) & ( 8.93 ) & ( 10.15 ) & ( 0.36 ) & ( 1.01 ) & ( 1.07 ) & ( 1.17 ) & ( 0.20 ) & ( 0.61 ) & ( 24.56 ) + + * [ -0.5ex] composite & 175 & 0.25 & 0.38 & 1.62 & 3.38 & 2.38 & 0.00 & 1.12 & 1.44 & 0.38 & 0.00 & 0.00 & 10.94 + & ( 157 ) & ( 0.56 ) & ( 0.51 ) & ( 1.07 ) & ( 3.60 ) & ( 0.76 ) & ( 0.00 ) & ( 0.81 ) & ( 0.61 ) & ( 0.05 ) & ( 0.00 ) & ( 0.00 ) & ( 7.97 ) + + * [ -0.5ex] seyfert & 40 & 0.19 & 0.31 & 0.25 & 0.56 & 0.25 & 0.00 & 0.44 & 0.38 & 0.12 & 0.00 & 0.00 & 2.50 + & ( 46 ) & ( 0.15 ) & ( 0.41 ) & ( 0.30 ) & ( 0.56 ) & ( 0.41 ) & ( 0.00 ) & ( 0.30 ) & ( 0.15 ) & ( 0.05 ) & ( 0.00 ) & ( 0.00 ) & ( 2.33 ) + + * [ -0.5ex] liner & 63 & 0.31 & 0.75 & 0.75 & 1.19 & 0.06 & 0.00 & 0.69 & 0.19 & 0.00 & 0.00 & 0.00 & 3.94 + & ( 107 ) & ( 2.44 ) & ( 0.61 ) & ( 1.01 ) & ( 0.61 ) & ( 0.10 ) & ( 0.00 ) & ( 0.30 ) & ( 0.36 ) & ( 0.00 ) & ( 0.00 ) & ( 0.00 ) & ( 5.43 ) + + * [ -0.5ex]weak - emission & 46 & 1.88 & 1.00 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 & 2.88 + & ( 229 ) & ( 7.76 ) & ( 3.40 ) & ( 0.30 ) & ( 0.05 ) & ( 0.00 ) & ( 0.00 ) & ( 0.05 ) & ( 0.05 ) & ( 0.00 ) & ( 0.00 ) & ( 0.00 ) & ( 11.62 ) + + * [ -0.5ex]unclear & 648 & 7.56 & 8.75 & 2.44 & 8.38 & 9.25 & 0.06 & 1.63 & 1.63 & 0.75 & 0.00 & 0.00 & 40.43 + & ( 948 ) & ( 17.96 ) & ( 13.25 ) & ( 4.42 ) & ( 5.88 ) & ( 3.19 ) & ( 0.00 ) & ( 2.08 ) & ( 1.07 ) & ( 0.25 ) & ( 0.00 ) & ( 0.00 ) & ( 48.10 ) + + * [ -0.5ex]total & 1605 & 10.19 & 11.56 & 6.00 & 26.44 & 29.37 & 0.81 & 4.44 & 5.88 & 4.00 & 0.12 & 1.19 & 100.00 + & ( 1973 ) & ( 28.92 ) & ( 18.32 ) & ( 7.97 ) & ( 19.63 ) & ( 14.61 ) & ( 0.36 ) & ( 4.57 ) & ( 3.30 ) & ( 1.52 ) & ( 0.20 ) & ( 0.61 ) & ( 100.00 ) [ tab : bpt ] ( * ? ? ? * herafter bpt ) demonstrated that several emission - line flux ratios are useful to classify emission line galaxies depending on their excitation mechanism . this diagnostic method has been revised by several studies @xcite .
we used the demarcation lines suggested by @xcite , @xcite and @xcite in order to categorize strong emission line galaxies into star forming , composite , low ionization nuclear emission - line region ( liner ) and seyfert types .
figure [ fig : bpt ] shows the bpt diagram of isolated galaxies ( upper panels ) and comparison sample galaxies ( lower panels ) with @xmath65 for all four emission lines ( [ ] @xmath665007 , h@xmath1 , [ ] @xmath666584 , and h@xmath2 ) . here
, @xmath67 is a ratio of the best fitting amplitude of emission line to residual noise of @xmath51 which is a similar to @xmath68 . in this figure
, the size and color of each symbol corresponds to the supermassive black hole mass and the ( g r)@xmath3 color of the host galaxy , respectively .
black hole mass is calculated using the m@xmath69@xmath26 relation in @xcite for galaxies with @xmath70 .
most of the galaxies with strong emission lines are late - type galaxies as shown in figure [ fig : bpt ] .
the contours of late - type galaxies ( right panels ) are predominantly found in the star forming region , and slightly extended to the composite region .
there is no significant difference between isolated and comparison galaxies .
furthermore , the trends that the black hole mass is larger in the agn region while color is bluer in the star forming region are same for isolated and comparison sample galaxies .
the distributions of early - type galaxies in the isolated and comparison samples on the bpt diagram seem different ( left - hand panels ) . for isolated galaxies , emission - line early - type galaxies
are more uniformly dispersed .
it is notable that a large fraction of emission - line early - type galaxies in the star forming and composite region are lenticular galaxies ( see the early - type galaxies in figure [ fig : bpt_fraction ] and table [ tab : bpt ] ) .
meanwhile , many of the emission - line early - type galaxies in the comparison sample are classified as liner , which is consistent with previous studies @xcite .
liners are generally considered to be agns @xcite , indicating a higher fraction of agns in the early - type galaxies of the comparison sample .
but we should bear in mind that there have been many debates on the excitation mechanism of liner features .
it may also be caused by the star formation @xcite , or old post - agb stars @xcite .
@xcite in particular demonstrated that because the agn activity is confined to the galactic center , other mechanisms such as star formation may affect the classification of `` true '' liner in the sdss , whose aperture size is larger than agn sources .
the fraction of the galaxies in each region is presented in figure [ fig : bpt_fraction ] .
liner , seyfert , composite and star forming region are the same as shown in figure [ fig : bpt ] and weak - emission galaxies are when @xmath71 for all 4 emission lines .
galaxies which are not weak - emission galaxies neither emission - line galaxies are classified as unclear - type galaxies and excluded in figure [ fig : bpt_fraction ] .
a considerable number of elliptical galaxies are weak - emission galaxies , and there is almost no weak - emission galaxies in late - type galaxies , as we would expect for star forming galaxies .
the different subclasses of elliptical galaxies do not show any significant trend , although the liner fractions in the comparison sample are slightly higher for all subclasses .
for late - type galaxies , the fraction of star forming galaxies increases from s(b)a to s(b)d , as they become increasingly late - type . meanwhile
the fraction in agn and composite region decreases along the hubble sequence for isolated and comparison sample galaxies .
the detailed information is summarized in table [ tab : bpt ] .
equivalent width of non - agn galaxies .
the format is the same as that of figure [ fig : cr].,scaledwidth=50.0% ] young stars ionize gas in starforming regions , forming regions . h@xmath2 emission is produced in these regions , and this has been studied frequently in the past ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
figure [ fig : h_alpha ] shows the h@xmath2 emission line strength in equivalent width for non - agn host galaxies which include star forming , composite , unclear - type and weak emission galaxies .
the median ew values of the h@xmath2 emission line are very small in early - type galaxies and there is no significant difference between isolated galaxies and comparison sample , because star forming early - type galaxies only account for a small fraction . in late - type galaxies , the strength of the h@xmath2 emission line increases from early to late - type spirals for both isolated and comparison galaxies .
@xcite and @xcite report that the h@xmath2 emission of interacting galaxies is stronger than that of an isolated galaxy .
in addition , @xcite claims that minor mergers can enhance the star formation in disk galaxies , and this enhancement is more significant in late - type spirals than in early - type spiral galaxies . in figure
[ fig : h_alpha ] , we show that the median values of h@xmath2 emission line strength of isolated galaxies are not significantly different from those of comparison galaxies for s(b)a and s(b)b galaxies . s(b)c and sd galaxies
( except for sbd galaxies due to the small number ) in the comparison sample may have slightly larger values than isolated ones , but the difference is not as conspicuous as reported in @xcite , perhaps partly due to our large error bars .
the most notable differences between the isolated and comparison samples can be summarized as follows : isolated galaxies are much more likely to be late type in morphology , and isolated early - type galaxies are 50% less massive and 20% younger than their counterparts . the most obvious candidate for the origin of such differences
may be found in the merger history of galaxies .
semi - analytic models can provide a comprehensive picture of galaxy evolution history according to different density environments ( e.g. , * ? ? ?
* ; * ? ? ?
in order to investigate the origin of the various galaxy properties in different environments , we use a semi - analytic model developed by @xcite based on a cosmological n - body volume simulation of structure formation run using gadget-2 @xcite . for the volume
, we adopt the cosmological parameters derived from the _ wilkinson microwave anisotropy probe seven year observations _
@xcite , @xmath72=0.272 , @xmath73 , and @xmath74 .
the periodic cube size of the volume is @xmath75mpc on a side with @xmath76 collision - less particles .
we generated a halo catalog by identifying substructures from friends - of - friends ( fof ) groups using subfind @xcite .
then , halo merger trees were constructed from the halo catalog by a tree building algorithm described in @xcite .
we used the merger trees as an input of the semi - analytic model .
further details of prescriptions for baryonic physics can be found in @xcite .
the results of the semi - analytic model are shown in figure [ fig : history ] . from left to right panels , the stellar mass is varied following @xmath77 ( panels a and e ) , @xmath78 ( b and f ) , @xmath79 ( c and g ) , and @xmath80 ( d and h ) and these stellar masses roughly correspond to the median values of the sc , sb , sa - type spiral and elliptical galaxies in our sample ( figure [ fig : mass ] ) .
also , the median bulge - to - total ratio of the galaxies in each panel is roughly 0.1 , 0.3 , 0.5 and 0.8 , respectively , which is roughly consistent with the observational data ( e.g. , * ? ? ?
the galaxies used in the semi - analytic model satisfy the density criteria of our sampling method considering the same stellar mass cut ( @xmath81 ) as the one used in our volume limitation on the sdss data .
upper panels of figure [ fig : history ] are the averaged evolution of the merger rate for isolated ( blue solid line ) and comparison galaxies ( red dashed line ) over the age of the universe . only mergers in a `` direct progenitor '' are taken into account when we calculate the merger rate . here
, a direct progenitor means the galaxy in the most massive halo in a merger tree at a given epoch@xcite .
also , only mergers with a stellar mass ratio greater than 1:10 ( which may significantly affect the galaxy s evolution ) are considered to derive the merger rate . in each panel ,
the mean merger rate is indicated in units of gyr@xmath4 per galaxy and the mean merger rates increase as galaxies become more massive , as expected .
interestingly , however , the mean merger rates ( the values in the top of each panel ) are similar within each panel ( that is , effectively for a given stellar mass ) regardless of the density of the environment . the merger - rate profiles of the two samples also seem similar to each other , perhaps with the exception of the first panel ( lowest mass sample ) .
spiral galaxies do not show much dependence on environment .
but this is sensible from the observational point of view
. spirals , whether in isolation or in dense regions , can not have had a significant number of recent major mergers , because otherwise they would not have remained as spirals to date .
this is supported by our model as well .
note that the merger rates for later type spirals ( e.g. , panel a ) are very low : typically only one important ( 1:10 or larger ) merger is expected for the last 10 billion years for these galaxies . as mentioned above ,
earlier type spiral galaxies have gradually higher merger rates ( top rows ) , and hence , if we consider all spiral galaxies together from sa to sd - type , the combined merger - rate profile of isolated galaxies will be lower than that of comparisons due to the high fraction of later - type spiral galaxies in isolated systems .
the similar merger rates of elliptical galaxies of the two samples ( panel d ) seem contrary to the expectation that the elliptical galaxies in denser regions may suffer significantly more mergers than isolated elliptical galaxies . however , the observed data suggest that isolated elliptical galaxies are 50% less massive than their counterparts in the comparison sample , and this must be considered in the comparison .
we present the profiles for such less massive elliptical galaxies from the models in panel ( d ) as gray solid lines .
the less massive elliptical galaxies had lower merger rates than more massive elliptical galaxies ( red and blue lines ) .
so , in this panel we should consider the red dashed line a representation for the elliptical galaxies in our comparison sample and the gray solid line for the isolated ellipticals .
sfrs normalized by the mean sfrs are shown in the lower panels .
the sfr presented here is the cumulative sfr of all progenitors ( galaxies in the host halo and subhalos ) .
the shape of the sfr evolution seems to follow that of the merger rate , which implies that mergers between galaxies play an important role in the star formation history in the framework of semi - analytic models . like the merger rate , the mean sfrs in units of m@xmath82/yr and the profiles of normalized sfrs are similar for isolated and comparison galaxies in the same mass range .
the luminosity - weighted age of a galaxy is sensitive to the star formation occurring during the past few gyr .
therefore , the similar normalized sfrs of isolated and comparison galaxies during the past 34 gyr ( hatched region ) can lead to the similar luminosity - weighted ages for a fixed stellar mass .
however , as mentioned above , the comparison galaxies in the highest mass bin should be compared with less massive isolated galaxies due to the difference in stellar mass . in panel
( h ) , the gray solid line ( isolated galaxies with @xmath83 ) shows slightly higher normalized sfr in recent time than the red dashed line ( comparison galaxies with @xmath84 ) .
this means that there are larger amounts of accumulated star formation in isolated galaxies during the last few gyr and therefore , the isolated elliptical galaxies would be measured to be younger than the comparison sample in luminosity - weighted age .
the results are compatible with those from the absorption - line analysis that isolated elliptical galaxies appear younger than their counterparts while spiral galaxies in the isolated and comparison samples have similar luminosity - weighted ages for a given morphology . in this interpretation , the tendency in the luminosity - weighted ages of galaxies , at least in the local universe , may be viewed as a galaxy mass effect rather than environmental effect @xcite . ) , b ( @xmath85 ) , c ( applying the cut in @xcite ) and d ( excluding the unknown type ) histograms are for different samples .
see the text for details.,scaledwidth=43.0% ] in this section , we discuss possible bias that could affect our works .
the situations where bias could be introduced are the following : case 1 : : the selection of the galaxies with @xmath85 from the volume - limited sample ( @xmath86 ) .
case 2 : : the use of the criteria in @xcite .
case 3 : : the classification of morphology into the unknown class during the visual inspection .
figure [ fig : bias ] shows the distribution of the stellar mass for isolated ( upper panel ) and comparison sample galaxies ( lower panel ) .
different histogram styles indicate different samples .
the dotted - line open histogram ( a ) is for the sample with @xmath86 .
the dashed - line open histogram ( b ) is for the sample with @xmath85 .
the filled histogram ( c ) is for the sample after applying the cut in @xcite .
the hatched histogram ( d ) is for the sample after excluding the unknown class through the visual inspection .
the overall distributions of the comparison sample are slightly shifted toward high mass compared to those of isolated galaxies . some low - mass isolated ( or comparison ) galaxies in the dotted - line open histogram ( a )
would be misclassified because of the inhomogeneous sampling ( see section [ sec : sample ] ) .
case 1 causes sample a to change to sample b , case 2 is from sample b to c , and case 3 is from sample c to d. since sd - type spiral galaxies have the lowest mass among spiral galaxies , they are likely to be abundant at @xmath87 @xmath88 . in case 1 ,
most galaxies with a mass range of @xmath14@xmath21@xmath89 @xmath88 are missed , so many sd - type galaxies would be removed , only leaving a small number of massive sd - type galaxies .
in addition , among sc - type galaxies , the galaxies with low mass would also be removed due to its median stellar mass being close to @xmath89 @xmath88 .
this would make sc - type spiral galaxies biased toward massive ones . in case 2 ,
the small and apparently flat galaxies are likely to be excluded .
thus , there is a chance that sb - type and sc - type spirals , as well as e6 and e7 elliptical galaxies , are missed . as discussed in section [ sec : morph ] , s0 galaxies are not easy to classify and thus , s0 galaxies and complex sa - type and elliptical galaxies would account for a large fraction of the unknown class in our study ( section [ sec : morph_compare ] ) .
as a result , some galaxies with the mass range of @xmath14@xmath89@xmath90 @xmath88 can be excluded in case 3 , and some late - type spiral galaxies are missed .
hence , through case 1 , 2 and 3 , spiral galaxies seem to be affected not only in number fraction but in stellar mass , whereas elliptical galaxies with mass @xmath91 @xmath88 are less affected . to summarize ,
although we have found that spiral galaxies are rich in isolated galaxies ( figure [ fig : fraction ] ) , the fraction of spiral galaxies would be underestimated due to the sampling bias .
the stellar mass in spiral galaxies could be somewhat biased toward massive galaxies in figure [ fig : mass ] , but this occurs to a lesser degree in elliptical galaxies .
we have used 1,644 isolated galaxies and 1,994 comparison galaxies , drawn from the sdss dr7 , in the redshift range of @xmath7 , to investigate the properties of isolated galaxies along the hubble morphological types .
our study focuses on not only photometric properties , but also spectroscopic characteristics of isolated galaxies , with the help of the ossy catalog which provides accurate line measurements .
the main results of this study can be summarized as follows . 1 . to study isolated galaxies , we construct the isolated galaxy sample and its comparison group based on the number of companion galaxies .
we define the galaxies without any companion galaxy within a 2 mpc sphere as isolated galaxies and those with at least two companions within 1 mpc as comparison sample galaxies . by choosing galaxies with @xmath85 from our volume - limited sample ,
we ensure we can detect companions down to a mass ratio of @xmath141:3 for our faintest galaxies .
this choice provides us with comparable numbers of galaxies between the isolated and comparison sample .
isolated galaxies are distributed widely and even located in high - density ( by long - range density measurement ) regions , whereas comparison galaxies are mainly positioned in high - density areas .
we have also found that there are a small number of void galaxies in our isolated galaxy sample as well as in the comparison set .
this implies that the isolated galaxies in our study are primarily influenced by the local density environment , rather than the global one .
morphologies of the two samples are classified through a visual inspection and the fractions along the hubble sequence of isolated and comparison galaxies show different distributions .
only 21.2% of isolated galaxies are early type whereas 46.7% are early type in the comparison sample . naturally , spiral galaxies are much more common in isolated systems ( 77.6% ) than in comparison ones ( 52.5% ) .
this result is consistent with the morphology - density relation @xcite .
subclasses ( e0e5 ) of elliptical galaxies in the two samples have similar distributions , because we do not consider the intrinsic shapes .
the fractions of barred spiral galaxies seem similar regardless of the density of the environment .
elliptical galaxies of both isolated and comparison galaxies form the red sequence on the color - magnitude diagram , but , on the whole , isolated ones ( @xmath45 ) appear fainter than their comparisons ( @xmath46 ) .
unbarred spiral galaxies of both samples are in the blue cloud and overall distributions are similar .
barred spiral galaxies show bimodal distributions without any distinct difference between the isolated and comparison samples .
5 . for elliptical galaxies
, there is a discrepancy in stellar mass between the isolated and comparison sample galaxies .
the median of the stellar mass in isolated elliptical galaxies ( @xmath50 ) is lower than that of their comparisons ( @xmath49 ) by 50% . on the other hand ,
spiral galaxies do not show any significant difference in their stellar mass between the two samples _ within subclasses_. the same trends are observed in the distributions of their velocity dispersion .
the median absorption line strengths including h@xmath1 , fe5270 and mgb between isolated galaxies and their comparisons are similar for all morphological types .
we used the tmj stellar population model @xcite which can help to break the degeneracy between the age and metallicity of a galaxy .
elliptical galaxies show a discrepancy between isolated and comparison galaxies in the median luminosity - weighted age by 20% and there is no obvious distinction in the median luminosity - weighted metallicity or [ @xmath2/fe ] .
spiral galaxies in the two samples do not show any significant difference in properties of stellar populations .
we have not found any significant difference in emission - line properties between isolated and comparison samples that can be explained in a straightforward way .
we constructed a sample of isolated galaxies following conventional approaches based on local density .
it is somewhat surprising that isolated galaxies do not show much difference in observable properties , but this is largely supported by the current hierarchical prediction .
it seems that the `` isolated '' galaxies defined this way are not an extreme population of galaxies after all , as long as they are part of the same large scale structures .
we hereby provide the public with the tables that contain the list of all of our sample galaxies and their properties measured from the observed data and derived from the comparison with stellar population models .
sky acknowledges support from the national research foundation of korea ( doyak 2014003730 ) .
numerical simulations were performed using the kisti supercomputer under the programme of ksc-2013-c3 - 015 .
this study was performed under the drc collaboration between yonsei university and the korea astronomy and space science institute . *
[ -0.5ex]% & 0.36 & 2.25 & 2.55 & 2.86 & 1.52 & 0.36 & 9.91 & 11.25 + & ( 1.25 ) & ( 7.57 ) & ( 8.83 ) & ( 6.97 ) & ( 2.91 ) & ( 1.05 ) & ( 28.59 ) & ( 18.10 ) + + * [ -0.5ex]cr & 3.21@xmath92 & 3.14@xmath93 & 3.18@xmath94 & 3.24@xmath94 & 3.24@xmath95 & 3.14@xmath96 & 3.20@xmath97 & 3.04@xmath98 + & ( 3.18@xmath99 ) & ( 3.19@xmath100 ) & ( 3.25@xmath101 ) & ( 3.23@xmath102 ) & ( 3.17@xmath103 ) & ( 3.22@xmath104 ) & ( 3.22@xmath105 ) & ( 3.07@xmath106 ) + + * [ -0.5ex]`fracdev ` & 0.98@xmath107 & 1.00@xmath108 & 1.00@xmath109 & 1.00@xmath110 & 1.00@xmath110 & 0.98@xmath111 & 1.00@xmath112 & 1.00@xmath113 + & ( 1.00@xmath108 ) & ( 1.00@xmath112 ) & ( 1.00@xmath114 ) & ( 1.00@xmath112 ) & ( 1.00@xmath115 ) & ( 1.00@xmath113 ) & ( 1.00@xmath112 ) & ( 1.00@xmath113 ) + + * [ -0.5ex]@xmath116 & 0.77@xmath117 & 0.76@xmath118 & 0.75@xmath119 & 0.76@xmath119 & 0.75@xmath118 & 0.79@xmath120 & 0.76@xmath119 & 0.75@xmath119 + & ( 0.77@xmath118 ) & ( 0.77@xmath121 ) & ( 0.77@xmath122 ) & ( 0.77@xmath122 ) & ( 0.77@xmath122 ) & ( 0.75@xmath121 ) & ( 0.77@xmath122 ) & ( 0.76@xmath122 ) + + * [ -0.5ex]@xmath47 & @xmath123 & @xmath124 & @xmath125 & @xmath126 & @xmath127 & @xmath128 & @xmath129 & @xmath130 + & ( @xmath131 ) & ( @xmath132 ) & ( @xmath133 ) & ( @xmath134 ) & ( @xmath135 ) & ( @xmath136 ) & ( @xmath137 ) & ( @xmath138 ) + + * [ -0.5ex]log(@xmath139/@xmath88 ) & 10.90@xmath140 & 10.96@xmath106 & 10.87@xmath141 & 10.92@xmath142 & 10.94@xmath143 & 10.99@xmath144 & 10.92@xmath145 & 10.76@xmath97 + & ( 11.12@xmath142 ) & ( 11.04@xmath146 ) & ( 11.14@xmath98 ) & ( 11.11@xmath147 ) & ( 11.12@xmath148 ) & ( 11.02@xmath97 ) & ( 11.10@xmath149 ) & ( 10.83@xmath150 ) * [ -0.5ex]% & 5.84 & 26.28 & 30.11 & 0.91 & 4.32 & 5.78 & 4.32 & 0.12 & 1.16 + & ( 7.87 ) & ( 19.66 ) & ( 15.25 ) & ( 0.40 ) & ( 4.51 ) & ( 3.26 ) & ( 1.55 ) & ( 0.20 ) & ( 0.60 ) + + * [ -0.5ex]cr & 2.73@xmath151 & 2.23@xmath152 & 2.11@xmath153 & 2.10@xmath154 & 2.46@xmath155 & 2.16@xmath141 & 2.10@xmath156 & 2.16@xmath157 & 2.61@xmath158 + & ( 2.71@xmath159 ) & ( 2.24@xmath160 ) & ( 2.13@xmath161 ) & ( 2.11@xmath162 ) & ( 2.48@xmath163 ) & ( 2.31@xmath164 ) & ( 2.15@xmath165 ) & ( 2.07@xmath166 ) & ( 2.33@xmath167 ) + + * [ -0.5ex]`fracdev ` & 0.84@xmath168 & 0.18@xmath169 & 0.12@xmath170 & 0.09@xmath171 & 0.83@xmath172 & 0.46@xmath173 & 0.26@xmath174 & 0.28@xmath157 & 0.35@xmath175 + & ( 0.82@xmath176 ) & ( 0.24@xmath177 ) & ( 0.15@xmath178 ) & ( 0.02@xmath179 ) & ( 0.89@xmath180 ) & ( 0.69@xmath181 ) & ( 0.36@xmath182 ) & ( 0.25@xmath183 ) & ( 0.26@xmath184 ) + + * [ -0.5ex]@xmath116 & 0.67@xmath185 & 0.56@xmath186 & 0.49@xmath187 & 0.39@xmath188 & 0.67@xmath189 & 0.60@xmath190 & 0.49@xmath191 & 0.37@xmath157 & 0.37@xmath192 + & ( 0.70@xmath193 ) & ( 0.57@xmath194 ) & ( 0.48@xmath195 ) & ( 0.33@xmath196 ) & ( 0.70@xmath189 ) & ( 0.61@xmath197 ) & ( 0.47@xmath198 ) & ( 0.36@xmath121 ) & ( 0.37@xmath199 ) + + * [ -0.5ex]@xmath47 & @xmath200 & @xmath201 & @xmath202 & @xmath203 & @xmath204 & @xmath205 & @xmath206 & @xmath207 & @xmath208 + & ( @xmath209 ) & ( @xmath210 ) & ( @xmath211 ) & ( @xmath212 ) & ( @xmath213 ) & ( @xmath214 ) & ( @xmath215 ) & ( @xmath216 ) & ( @xmath217 ) + + * [ -0.5ex]log(@xmath139/@xmath88 ) & 10.84@xmath218 & 10.48@xmath219 & 10.34@xmath220 & 10.08@xmath221 & 10.80@xmath222 & 10.64@xmath223 & 10.28@xmath224 & 10.09@xmath157 & 10.07@xmath225 + & ( 10.89@xmath226 ) & ( 10.55@xmath227 ) & ( 10.35@xmath228 ) & ( 9.96@xmath229 ) & ( 10.88@xmath230 ) & ( 10.74@xmath231 ) & ( 10.15@xmath232 ) & ( 10.14@xmath233 ) & ( 10.04@xmath234 ) * [ -0.5ex]@xmath235 ( ) & 127.6@xmath236 & 169.3@xmath237 & 149.1@xmath238 & 175.3@xmath239 & 167.1@xmath240 & 174.9@xmath241 & 160.6@xmath242 & 133.6@xmath243 + & ( 198.1@xmath244 ) & ( 187.0@xmath245 ) & ( 201.5@xmath246 ) & ( 196.1@xmath247 ) & ( 194.6@xmath248 ) & ( 183.1@xmath249 ) & ( 194.2@xmath250 ) & ( 143.6@xmath251 ) + + * [ -0.5ex]h@xmath1 ( ) & 2.02@xmath252 & 1.84@xmath253 & 1.82@xmath141 & 1.76@xmath254 & 1.89@xmath255 & 1.87@xmath256 & 1.83@xmath59 & 1.90@xmath257 + & ( 1.76@xmath258 ) & ( 1.74@xmath259 ) & ( 1.78@xmath260 ) & ( 1.72@xmath261 ) & ( 1.68@xmath178 ) & ( 1.73@xmath99 ) & ( 1.75@xmath60 ) & ( 1.88@xmath262 ) + + * [ -0.5ex]fe5270 ( ) & 2.79@xmath263 & 2.84@xmath264 & 2.72@xmath265 & 2.81@xmath266 & 2.76@xmath267 & 2.92@xmath268 & 2.80@xmath269 & 2.80@xmath220 + & ( 2.93@xmath270 ) & ( 2.83@xmath271 ) & ( 2.87@xmath272 ) & ( 2.84@xmath273 ) & ( 2.84@xmath274 ) & ( 2.82@xmath275 ) & ( 2.85@xmath276 ) & ( 2.86@xmath272 ) + + * [ -0.5ex]@xmath30fe@xmath64 ( ) & 2.76@xmath277 & 2.84@xmath278 & 2.73@xmath148 & 2.81@xmath279 & 2.70@xmath280 & 2.95@xmath281 & 2.80@xmath98 & 2.80@xmath220 + & ( 2.92@xmath282 ) & ( 2.83@xmath271 ) & ( 2.87@xmath147 ) & ( 2.84@xmath273 ) & ( 2.85@xmath95 ) & ( 2.87@xmath283 ) & ( 2.85@xmath276 ) & ( 2.86@xmath272 ) + + * [ -0.5ex]@xmath284mgfe@xmath285 $ ] ( ) & 3.12@xmath286 & 3.34@xmath287 & 3.25@xmath275 & 3.35@xmath288 & 3.20@xmath228 & 3.45@xmath289 & 3.28@xmath290 & 3.24@xmath291 + & ( 3.49@xmath292 ) & ( 3.43@xmath279 ) & ( 3.47@xmath293 ) & ( 3.47@xmath294 ) & ( 3.50@xmath295 ) & ( 3.46@xmath296 ) & ( 3.47@xmath297 ) & ( 3.35@xmath298 ) + + * [ -0.5ex]mgb ( ) & 3.85@xmath299 & 4.03@xmath300 & 3.91@xmath301 & 4.09@xmath302 & 3.81@xmath303 & 3.86@xmath304 & 4.02@xmath61 & 3.83@xmath305 + & ( 4.33@xmath306 ) & ( 4.26@xmath307 ) & ( 4.32@xmath308 ) & ( 4.33@xmath309 ) & ( 4.40@xmath310 ) & ( 4.29@xmath311 ) & ( 4.32@xmath62 ) & ( 4.02@xmath312 ) + + * [ -0.5ex]log(@xmath313/@xmath88 ) & 7.29@xmath314 & 7.81@xmath315 & 7.58@xmath316 & 7.88@xmath317 & 7.79@xmath318 & 7.87@xmath319 & 7.71@xmath320 & 7.38@xmath321 + & ( 8.10@xmath322 ) & ( 8.00@xmath323 ) & ( 8.13@xmath324 ) & ( 8.08@xmath325 ) & ( 8.07@xmath326 ) & ( 7.96@xmath327 ) & ( 8.07@xmath328 ) & ( 7.51@xmath329 ) + + * [ -0.5ex]h@xmath2 ( ) & 0.52@xmath330 & 0.48@xmath331 & 0.61@xmath332 & 0.48@xmath333 & 0.41@xmath334 & 0.81@xmath335 & 0.49@xmath336 & 0.49@xmath337 + & ( 0.33@xmath338 ) & ( 0.35@xmath339 ) & ( 0.35@xmath340 ) & ( 0.38@xmath341 ) & ( 0.53@xmath342 ) & ( 0.63@xmath343 ) & ( 0.39@xmath344 ) & ( 0.45@xmath345 ) * [ -0.5ex]@xmath235 ( ) & 114.0@xmath346 & 61.9@xmath347 & 48.8@xmath348 & 40.4@xmath349 & 101.8@xmath350 & 75.5@xmath351 & 53.2@xmath352 & 0.0@xmath353 & 51.1@xmath354 + & ( 122.3@xmath355 ) & ( 69.8@xmath356 ) & ( 52.1@xmath357 ) & ( 57.2@xmath358 ) & ( 105.4@xmath359 ) & ( 92.1@xmath360 ) & ( 53.6@xmath361 ) & ( 38.1@xmath353 ) & ( 42.2@xmath362 ) + + * [ -0.5ex]h@xmath1 ( ) & 2.25@xmath363 & 2.90@xmath364 & 2.93@xmath365 & 3.98@xmath366 & 2.37@xmath367 & 2.64@xmath368 & 3.13@xmath369 & 4.41@xmath157 & 4.01@xmath155 + & ( 2.11@xmath363 ) & ( 2.84@xmath370 ) & ( 3.14@xmath371 ) & ( 4.14@xmath372 ) & ( 2.16@xmath373 ) & ( 2.43@xmath374 ) & ( 3.14@xmath375 ) & ( 4.33@xmath376 ) & ( 4.18@xmath377 ) + + * [ -0.5ex]fe5270 ( ) & 2.56@xmath328 & 2.14@xmath378 & 1.95@xmath379 & 1.06@xmath380 & 2.48@xmath381 & 2.31@xmath382 & 1.79@xmath383 & 1.12@xmath157 & 1.24@xmath384 + & ( 2.64@xmath385 ) & ( 2.12@xmath386 ) & ( 1.89@xmath387 ) & ( 1.07@xmath388 ) & ( 2.61@xmath389 ) & ( 2.46@xmath390 ) & ( 1.77@xmath391 ) & ( 1.29@xmath392 ) & ( 1.06@xmath393 ) + + * [ -0.5ex]@xmath30fe@xmath64 ( ) & 2.56@xmath394 & 2.18@xmath395 & 2.02@xmath396 & 1.01@xmath397 & 2.49@xmath398 & 2.42@xmath399 & 1.98@xmath400 & 0.50@xmath157 & 1.24@xmath401 + & ( 2.67@xmath402 ) & ( 2.16@xmath403 ) & ( 1.94@xmath404 ) & ( 0.84@xmath405 ) & ( 2.62@xmath406 ) & ( 2.46@xmath407 ) & ( 2.10@xmath408 ) & ( 0.83@xmath157 ) & ( 0.84@xmath409 ) + + * [ -0.5ex]@xmath284mgfe@xmath285$]( ) & 2.82@xmath410 & 2.18@xmath411 & 2.03@xmath412 & 1.20@xmath413 & 2.74@xmath414 & 2.52@xmath415 & 2.03@xmath416 & 0.50@xmath157 & 1.22@xmath417 + & ( 3.03@xmath418 ) & ( 2.20@xmath419 ) & ( 1.89@xmath420 ) & ( 0.61@xmath274 ) & ( 2.97@xmath421 ) & ( 2.63@xmath422 ) & ( 2.07@xmath423 ) & ( 0.88@xmath157 ) & ( 0.92@xmath424 ) + + * [ -0.5ex]mgb ( ) & 3.17@xmath425 & 2.21@xmath426 & 2.03@xmath427 & 1.19@xmath290 & 3.07@xmath428 & 2.59@xmath429 & 1.66@xmath430 & 0.75@xmath157 & 1.16@xmath431 + & ( 3.42@xmath432 ) & ( 2.29@xmath433 ) & ( 1.83@xmath434 ) & ( 0.70@xmath435 ) & ( 3.27@xmath436 ) & ( 2.74@xmath437 ) & ( 1.55@xmath438 ) & ( 1.03@xmath439 ) & ( 0.91@xmath440 ) + + * [ -0.5ex]log(@xmath313/@xmath88 ) & 7.10@xmath441 & 5.86@xmath442 & 5.25@xmath443 & 4.57@xmath444 & 6.87@xmath445 & 6.25@xmath446 & 5.06@xmath447 & 4.64@xmath157 & 5.22@xmath448 + & ( 7.20@xmath449 ) & ( 6.02@xmath450 ) & ( 5.30@xmath451 ) & ( 4.68@xmath452 ) & ( 6.94@xmath453 ) & ( 6.59@xmath454 ) & ( 4.49@xmath455 ) & ( 4.37@xmath456 ) & ( 4.36@xmath457 ) + + * [ -0.5ex]h@xmath2 ( ) & 4.44@xmath458 & 12.36@xmath459 & 12.50@xmath460 & 27.72@xmath461 & 5.87@xmath462 & 13.57@xmath463 & 19.95@xmath464 & 55.04@xmath465 & 59.09@xmath466 + & ( 1.66@xmath467 ) & ( 12.29@xmath468 ) & ( 16.53@xmath469 ) & ( 54.98@xmath470 ) & ( 2.11@xmath471 ) & ( 11.27@xmath472 ) & ( 23.21@xmath473 ) & ( 38.15@xmath474 ) & ( 47.28@xmath475 ) cccccc 587739828207353983 & 13 & 228.61371 & 20.78934 & 0.0397760 & 0.780 + 588017703493500993 & 21 & 240.40990 & 7.10945 & 0.0400339 & 0.723 + 587726100949434515 & 16 & 220.11459 & 3.68553 & 0.0279422 & 0.713 + 587739407337455890 & 23 & 232.81613 & 25.31393 & 0.0341272 & 0.490 + 587739829271855250 & 23 & 206.47068 & 26.77517 & 0.0298094 & 0.521 + [ 2pt ] @xmath47 & @xmath476 ( arcsec ) & @xmath477 ( arcsec ) & @xmath478 & ` fracdev ` & log(@xmath139/@xmath88 ) + [ 2pt ] @xmath479 & 63.033 & 46.404 & 3.321 & 1.000 & 11.290 + @xmath480 & 38.271 & 26.423 & 3.013 & 0.983 & 10.834 + @xmath481 & 66.449 & 24.693 & 2.598 & 0.610 & 10.525 + @xmath482 & 44.662 & 34.365 & 2.134 & 0.105 & 10.189 + @xmath483 & 66.997 & 32.829 & 2.082 & 0.000 & 10.558 + [ 2pt ] @xmath235 ( ) & @xmath484 ) ( ) & h@xmath1 ( ) & fe5270 ( ) & @xmath30fe@xmath64 ( ) & mgb ( ) + [ 2pt ] 293.879 & 3.362 & 1.658 & 3.045 & 3.005 & 5.072 + 109.202 & 2.331 & 2.233 & 2.955 & 2.656 & 3.312 + 76.974 & 3.516 & 2.265 & 2.880 & 2.688 &
2.608 + 33.622 & 9.431 & 2.217 & 1.427 & 1.243 & 2.696 + 59.220 & 4.451 & 2.503 & 1.663 & 1.313 & 2.471 + [ 2pt ] @xmath284mgfe@xmath285 $ ] ( ) & age ( gyr ) & [ z / h ] & [ @xmath2/fe ] & bpt class & log(@xmath313/@xmath88 ) + [ 2pt ] 3.915 & 10.789 & 0.321 & 0.363 & 4 & 8.829 + 3.038 & 3.111 & 0.262 & 0.328 & 5 & 7.006 + 2.689 & 3.062 & 0.097 & 0.120 & 5 & 6.362 + 1.889 & 8.056 & @xmath485 & @xmath486 & 5 & 4.836 + 1.904 & 4.704 & @xmath487 & @xmath488 & 1 & 5.879 + | isolated galaxies in low - density regions are significant in the sense that they are least affected by the hierarchical pattern of galaxy growth , and interactions with perturbers , at least for the last few gyr . to form a comprehensive picture of the star formation history of isolated galaxies
, we constructed a catalog of isolated galaxies and their comparison sample in relatively denser environments .
the galaxies are drawn from the sdss dr7 in the redshift range of @xmath0 .
we performed a visual inspection and classified their morphology following the hubble classification scheme . for the spectroscopic study , we make use of the ossy catalog .
we confirm most of the earlier understanding on isolated galaxies .
the most remarkable additional results are as follows .
isolated galaxies are dominantly late type with the morphology distribution ( e : s0 : s : irr ) = ( 9.9 : 11.3 : 77.6 : 1.2)% . the frequency of elliptical galaxies among isolated galaxies is only a third of that of the comparison sample .
most of the photometric and spectroscopic properties are surprisingly similar between isolated and comparison samples . however , early - type isolated galaxies are less massive by 50% and younger ( by h@xmath1 ) by 20% than their counterparts in the comparison sample .
this can be explained as a result of different merger and star formation histories for differing environments in the hierarchical merger paradigm .
we provide an on - line catalog for the list and properties of our sample galaxies . |
following ozawa @xcite , the * representativity * @xmath4 of a link @xmath5 is : @xmath6 where @xmath7 is the set of positive genus , closed surfaces @xmath1 containing @xmath0 , and a `` closed surface '' is compact and connected without boundary ; @xmath8 is the set of compressing disks for @xmath2 in @xmath9 ; and @xmath10 is the number of connected components ( i.e. points ) of @xmath11 .
this notion extends the earlier notion of representativity from graph theory . in 2011
@xcite , pardon applied representativity ( although he did not use this term ) to answer a question posed by gromov @xcite in 1983 regarding knot distortion .
the distortion of an embedded circle @xmath12 in @xmath13 is defined to be : @xmath14 where @xmath15 is arclength along @xmath12 , and @xmath16 is euclidean distance in @xmath13 .
gromov asked whether there exists a uniform upper bound on distortion for all isotopy classes of knots , or at least for torus knots .
specifically , gromov asked , does every isotopy class of knots have a representative with distortion less than , say , 100 ? to answer this question , pardon showed that every knot isotopy class @xmath17 satisfies : @xmath18 where @xmath19 denotes representativity .
in particular , since the representativity of any @xmath20 torus knot is @xmath21 ( more to the point and easier to check is @xmath22 ) , so that @xmath23 as @xmath24 , pardon was able to answer gromov s question in the negative .
current work of blair , campisi , taylor and tomova @xcite provides a lower bound for @xmath25 in terms of distance and bridge numbers .
our main result implies that their lower bound improves pardon s lower bound in the case of alternating knots with sufficiently large distance : every non - trivial , non - split alternating link @xmath0 has representativity @xmath26 . as another application
, the main theorem gives a new proof of the fact that the only alternating torus links are the 2-braids @xmath27 , again since @xmath28 .
as @xmath29 is almost - alternating , the main theorem does not extend to this class of links .
we return to this example in [ s:5 ] , suggesting a possible approach to the characterization of almost - alternating links with representativity @xmath30 .
to prove the main theorem , we employ the crossing ball structures introduced by william menasco @xcite .
roughly , the idea is to insert a ball @xmath31 at each crossing of a given diagram @xmath32 on @xmath33 , to perturb @xmath0 to lie on @xmath34 , where @xmath35 , and then to isotope a given closed surface @xmath36 ( fixing @xmath0 and the crossing ball structure @xmath37 ) so as to minimize its intersections with @xmath38 and @xmath33 away from @xmath0 .
we show that whenever @xmath2 is essential ( incompressible and @xmath3-incompressible in the link exterior @xmath39 ) , there exists an isotopy of @xmath2 which produces a standard tube near some crossing ( cf .
figure [ fi : smoothing ] ) . given a non - trivial
, connected , reduced alternating diagram of a link @xmath0 and a closed , essential surface @xmath36 , there exists an isotopy of @xmath2 after which some crossing has a standard * tube*. this lemma not only provides a compressing disk for @xmath2 whose boundary intersects @xmath0 in two points ; it also provides a possible inductive move , in the tradition of @xcite , albeit one still awaiting application .
* thank you * to seungwon kim for sharing this problem during a visit to iowa ; to colin adams for introducing the author to menasco s techniques ( in particular to `` bigon moves '' ) ; to maggy tomova and ryan blair for helpful discussions around early and final versions of this paper , respectively ; and to charlie frohman for patient and inspired coaching .
a link diagram @xmath40 is the image of an immersion of one or more circles in @xmath41 in which all self - intersections are double - points at which the two intersecting arcs are transverse in @xmath33 and are labeled with opposite normal directions relative to @xmath33 .
thus , a link diagram @xmath32 can also be seen as a smoothly embedded 4-valent graph in @xmath33 with over - under information at each vertex .
an embedding of the underlying link @xmath0 can be obtained by perturbing the two arcs of @xmath32 near each crossing point in the indicated normal directions .
a link diagram @xmath32 is called * alternating * if , for each edge of @xmath32 ( seen as a 4-valent graph ) , the crossing points at its two ends are labeled with opposite normal directions . a link is called alternating if it has an alternating diagram . a link diagram @xmath32
is called * reduced * if it lacks _ nugatory _ crossings , which means that every crossing point is incident to four distinct components of @xmath42 .
a link @xmath0 is called * split * if @xmath43 is reducible , i.e. if there is an embedded 2-sphere in @xmath43 which does not bound a ball in @xmath43 .
let @xmath32 be a diagram of a link @xmath0 ( in [ s:3]-[s:4 ] , we will assume further that @xmath32 is non - trivial , connected , reduced , and alternating ) with crossing points @xmath44 , @xmath45 . insert small , mutually disjoint ( closed ) * crossing balls * @xmath31 , @xmath45 , centered at the respective crossing points @xmath44 .
denote @xmath46 .
construct an embedding of @xmath0 in @xmath47 by perturbing the two arcs in which @xmath32 intersects each crossing ball @xmath31 in their indicated normal directions from @xmath48 to @xmath49 , while fixing @xmath50 . call each resulting component of @xmath51 an * edge * of @xmath0note that @xmath52and call each component of @xmath53 an * overpass * or an * underpass * of @xmath0 , according to which side of @xmath33 it lies on . near each crossing , this looks like figure [ fi : crossingball ] ( center ) .
( left ) and @xmath54 ( right).,width=624 ] let @xmath56 be a closed regular neighborhood of the link @xmath0 , viewed as ( the total space of ) a disk - bundle @xmath57 for which the restrictions @xmath58 , @xmath59 are also bundle maps onto their images ( each with fiber a closed interval ) .
thus , each component of @xmath55 intersects @xmath47 in an annular neighborhood of its core ; and , for each point @xmath60 , the boundary of the disk @xmath61 is a meridian on @xmath62 which consists of an arc of @xmath63 glued to an arc of @xmath64 at the two points of @xmath65 .
let @xmath66 and @xmath67 be the two components into which @xmath68 cuts @xmath9 ; i.e. , @xmath69 are the closures of the components of @xmath70 .
let @xmath71 and @xmath72 .
near each crossing , @xmath73 and @xmath54 appear as in figure [ fi : crossingball ] ( left and right ) . as a quick exercise , check that @xmath74 and @xmath75 .
use this setup to extend the terminology of edges , overpasses , and underpasses from @xmath0 to @xmath62 as follows . call each component of @xmath76 an * overpass * of @xmath62 ,
call each component of @xmath77 an * underpass * of @xmath62 , and call each component of @xmath78 an * edge - pair * ; each edge - pair is the union of two * edges * of @xmath62 , one a component of @xmath79 , the other a component of @xmath80 .
this terminology gives @xmath62 the following cell decomposition ( cf .
figures [ fi : crossingball ] , [ fi : edgeblank ] ) . * the 0-cells are the points of @xmath81 , eight on the boundary of each crossing ball .
* there are several types of ( closures of ) 1-cells : * * arcs of @xmath82 , two running along each edge - pair _ of @xmath62 _ ; * * arcs of @xmath83 , two along each overpass of @xmath62 ; * * arcs of @xmath84 , two along each underpass of @xmath62 ; and * * arcs of @xmath85 , eight near each crossing ball @xmath31 : * * * four in @xmath86 , * * * two in @xmath87 , joining the overpass of @xmath62 at @xmath31 with edges of @xmath62 , and * * * two in @xmath88 , joining the underpass of @xmath62 at @xmath31 with edges of @xmath62 . *
the 2-cells closures are overpasses , underpasses , and edges of @xmath62 ; plus components of @xmath89 . from an alternating link diagram .
, title="fig:",height=134 ] from an alternating link diagram .
, title="fig:",height=163 ] so far , the link @xmath90 follows a link diagram @xmath40 ; @xmath55 is a closed regular neighborhood of @xmath0 in @xmath9 , seen as ( the total space of ) a disk - bundle @xmath57 ; and @xmath69 are the components into which @xmath68 cuts @xmath9 , with @xmath91
. now also let @xmath2 be any closed surface in @xmath9 of positive genus that contains @xmath0 .
( recall that a `` closed surface '' is assumed to be compact and connected without boundary ) .
fixing @xmath92 , @xmath33 , and @xmath38 , isotope @xmath2 so that : * @xmath2 is transverse to @xmath73 and @xmath54 ; * the restriction @xmath93 is a bundle map ( so that @xmath94 is a regular neighborhood of @xmath0 in @xmath2 and each component of @xmath95 is the image of a section of @xmath57 ) ; and * @xmath96 ( so that @xmath2 intersects each overpass and underpass of @xmath62 in precisely two arcs ) .
perform such isotopy so as to * minimize lexicographically * the numbers of components of @xmath97 and of @xmath98 , i.e. to minimize the * complexity * @xmath99 , where bars count connected components .
note that the first and last conditions above ensure that @xmath97 will consist only of simple closed curves , since the endpoints of any arc of @xmath97 would lie on @xmath100 .
the surface @xmath2 is said to be * compressible * in the link exterior @xmath101 if there is a disk @xmath102 with @xmath103 a simple closed curve ( a `` circle '' ) that does not bound a disk in @xmath2 .
the surface @xmath2 is compressible in @xmath104 if and only if @xmath105 .
in particular , every split link @xmath0 has representativity @xmath106 . the surface @xmath2 is said to be * @xmath3-compressible * in the link exterior @xmath104 if there is an arc @xmath107 on @xmath62 which is parallel to @xmath2 through @xmath104say , through a disk @xmath108 , with @xmath109 denoting the arc @xmath110but not through @xmath62 .
in this case , @xmath107 is also parallel through a bigon @xmath111 to an arc @xmath112 which intersects @xmath0 in one point . gluing @xmath108 and @xmath113 along @xmath107 produces a compressing disk @xmath114 for @xmath2 in @xmath9 whose boundary @xmath115 intersects @xmath0 in one point . thus , if @xmath2 is @xmath3-compressible in @xmath104 , then @xmath116 ; conversely as well .
in particular , the trivial knot @xmath0 has representativity @xmath117 , since any positive genus , closed surface containing @xmath0 must be compressible or @xmath3-compressible .
a closed surface @xmath36 is called * essential * if it is neither compressible nor @xmath3-compressible in the link exterior . any non - trivial _ knot _
@xmath0 is contained in a positive genus , closed surface @xmath2 with @xmath118 , namely the interpolating surface for any incompressible seifert surface for @xmath0 , or more generally for any algebraically essential spanning surface for @xmath0 .
( a spanning surface @xmath119 for @xmath0 is an embedded surface with boundary @xmath120 in @xmath9 ; the interpolating surface for a spanning surface @xmath119 is the boundary of a regular neighborhood of @xmath119 in the link exterior ; and @xmath119 is called algebraically essential if its interpolating surface is incompressible and @xmath3-incompressible in the link exterior . ) is this also true of non - split _ links _
@xmath0 ? is @xmath121 _ if _ and only if @xmath0 is non - split and non - trivial ? this is true for all non - split , non - trivial links with an algebraically essential , _ connected _ spanning surface , using the interpolating surface . does every non - split link have such a span ? at least in the alternating case , the answer is yes ( in fact , _ all _ spanning surfaces for non - split alternating links are connected @xcite ) .
thus , an alternating link @xmath0 in a reduced alternating diagram @xmath32 obeys : @xmath122 the main theorem states that whenever @xmath0 is alternating , non - trivial , and non - split , @xmath26 .
the three results in [ s:24 ] assume the following setup from [ s:21]-[s:23 ] .
a link @xmath90 follows a non - trivial , connected link diagram @xmath40 with @xmath123 crossings , and @xmath55 is a regular neighborhood of @xmath0 , seen as ( the total space ) of a disk - bundle @xmath57 .
balls @xmath69 are the closures of the two components of @xmath70 , with @xmath91 and @xmath75 . and a closed essential surface @xmath36 has been isotoped , subject to the conditions that @xmath2 is transverse to @xmath73 and @xmath54 , @xmath93 is a bundle map , and @xmath96 , so as to minimize its complexity @xmath124 .
this setup implies : [ p : disks ] all components of @xmath125 are disks .
the @xmath126 components of @xmath127namely @xmath66 , @xmath67 , and @xmath128 , @xmath45are all topological 3-balls , and their boundaries are the 2-spheres @xmath73 , @xmath54 , and @xmath129 .
these spheres intersect @xmath2 transversally , hence in circles .
each such circle bounds a disk in the corresponding ball .
the incompressibility of @xmath2 implies that each of these circles also bounds a disk in @xmath2 .
minimality then implies that this disk in @xmath2 must lie entirely in the appropriate ball , as claimed .
specifically , each circle of @xmath130 bounds a disk of @xmath131 , and each circle of @xmath132 bounds a disk of @xmath133 .
throughout [ s : broader ] , maintain all setup from [ s:21]-[s:23 ] , but replace the assumption that the complexity of @xmath2 is minimized with the assumption that all components of @xmath125 are disks .
proposition [ p : disks ] implies that this setting is more general than the initial setup .
[ p : arcs ] all components of @xmath134 , @xmath135 , and @xmath98 are arcs .
every component of @xmath55 contains an overpass and an underpass , or else @xmath0 would be a split link , and @xmath2 ( assumed to be connected ) would be compressible , contrary to assumption .
therefore , each component of @xmath95 , ( the image of ) a section of @xmath57 by assumption , must intersect @xmath136 .
hence , no component of @xmath134 is a circle ; instead , each must be an arc .
all components of @xmath137 , @xmath138 , and @xmath136 are disks . if @xmath2 intersected one of these disks in a circle , @xmath12 , then all components of @xmath125 are disks , @xmath12 would bound disks of @xmath2 in both components of @xmath125 whose boundaries contain @xmath12 .
this contradicts the assumption that @xmath2 is connected .
the conclusion of proposition [ p : arcs ] , implies that the number of components of @xmath98 equals half the number of points of @xmath139 , which will be more convenient to count .
also , the conclusions of propositions [ p : disks ] and [ p : arcs ] together imply that @xmath140 has the following cell decomposition : * the 0-cells are the points of @xmath141 .
* there are three types of ( closures of ) 1-cells : arcs of @xmath98 , @xmath135 , and @xmath134 . * the ( closures of the ) 2-cells are the components of @xmath142 , @xmath143 , and @xmath144 .
[ p : balls ] all components of @xmath145 are 3-balls .
let @xmath146 be a component of @xmath145 , and let @xmath108 be a component of @xmath147 .
since @xmath2 is connected , @xmath108 must contain ( at least ) one component @xmath113 of @xmath148 .
the boundary of @xmath113 is a union of circles of @xmath130 , each of which bounds a disk of @xmath131 , by assumption .
the union of these disks and @xmath113 is a closed surface and therefore equals @xmath108 ; computing euler characteristic reveals that @xmath108 is a 2-sphere in @xmath69 and therefore bounds a 3-ball in @xmath69 .
this 3-ball must be @xmath146 ; otherwise , since each component of @xmath147 is a 2-sphere which bounds a 3-ball in @xmath69 , gluing each of these 3-balls to @xmath146 along @xmath147 would yield a closed 3-manifold ( compact and connected without boundary ) contained in @xmath69 .
.,width=624 ] past a parallel arc @xmath149 , provided @xmath107 is not parallel to @xmath2 through @xmath150 and @xmath151.,width=624 ] several proofs in [ s:3]-[s:4 ] will use an isotopy move which pushes an arc @xmath152 past a parallel arc @xmath153 through a disk @xmath114 with @xmath154 .
one type of this `` bigon move '' is illustrated in figure [ fi : bigon ] ; all three types are diagrammed in figure [ fi : move ] .
more precisely , a bigon move follows an arc @xmath155 that : * intersects @xmath2 precisely on its endpoints , which lie on the same circle @xmath12 of @xmath130 ; * is not parallel in @xmath150 to @xmath2 ; * is disjoint from @xmath156 , i.e. from crossing balls and over / underpasses ; and * intersects @xmath136 in exactly one component .
think of @xmath107 , which initially is not `` part of the diagram , '' as a marker which joins two points that lie on the same circle @xmath12 of @xmath130 , but `` not obviously '' so , locally . because the circle @xmath12 bounds a disk @xmath157 , there is an arc @xmath158 with the same endpoints as @xmath107 .
the circle @xmath159 lies on the boundary of some component of @xmath145 , a ball , and therefore bounds a disk @xmath160 whose interior is disjoint from @xmath2 .
the disk @xmath114 is a bigon in the sense that @xmath161 .
[ bigonmove ] given an arc @xmath155 that satisfies the four conditions above , it is possible to isotope @xmath2 near a parallel arc @xmath152 through a bigon @xmath114 past @xmath107 .
all bigon moves take place away from the crossing balls , and thus preserve both the number of components of @xmath97 and the fact that each of these components bounds a disk of @xmath162 .
not all bigon moves , however , preserve the fact that all components of @xmath163 are disks .
at least : [ l : move ] performing a bigon move preserves the fact that all components of @xmath163 are disks whenever @xmath164 and whenever the complexity of @xmath2 is minimized .
as noted , any bigon moves preserves the fact that all components of @xmath144 are disks . in order for a bigon move to upset the fact that all components of @xmath142 and @xmath143 are disks
, a necessary condition is that the endpoints of @xmath107 lie on the same circle of @xmath130 and also on the same circle of @xmath165 . for this to be the case ,
both endpoints of @xmath107 must lie in @xmath136 , as must all of @xmath107 , since @xmath166 is connected by assumption .
thus , the condition @xmath164 in the statement of the lemma suffices as claimed .
suppose instead that , with the complexity of @xmath2 minimized , a bigon move follows an arc @xmath167 whose endpoints lie both on the same circle @xmath168 of @xmath130 _ and _ on the same circle @xmath169 of @xmath165 . letting @xmath113 , @xmath170 denote the disks of @xmath142 ,
@xmath143 respectively bounded by @xmath12 , @xmath171 , there are arcs @xmath158 , @xmath172 with the same endpoints as @xmath107 .
further , @xmath159 , @xmath173 respectively bound disks @xmath160 , @xmath174 whose interiors are disjoint from @xmath2 .
gluing @xmath114 and @xmath175 along @xmath107 produces a disk @xmath176 with boundary @xmath177 whose interior is disjoint from @xmath2 ; since @xmath2 is incompressible in the link exterior , @xmath178 must bound a disk @xmath179 . since @xmath0 is non - split , the 2-sphere @xmath180 bounds a 3-ball @xmath181 , through which @xmath108 is parallel to @xmath176 . since the complexity of @xmath2 is minimized , the disk @xmath108 , like @xmath176 , must be disjoint from @xmath38 and must intersect @xmath136 in a single arc , @xmath182 .
from @xmath183 , it follows that @xmath184 .
this and the fact that @xmath185 imply that @xmath186 is disjoint from @xmath187 , and in particular from @xmath188 .
therefore , contrary to assumption , @xmath186 intersects @xmath136 in a single disk through which @xmath107 is parallel to the arc @xmath189 : @xmath190 note that a bigon move fixes the number of components of @xmath98which equals half the number of points of @xmath191if and only if @xmath192 ; otherwise , a bigon move increases this number , and thus the complexity of @xmath2 .
in particular : [ l : movecase ] if sequence of bigon moves begins with the complexity of @xmath2 minimized , and if all bigon moves in this sequence along arcs disjoint from @xmath62 precede all other bigon moves in this sequence , then this sequence of bigon moves preserves the fact that all components of @xmath193 are disks . because bigon moves along arcs disjoint from @xmath62 fix the complexity of @xmath2 , all such moves in this sequence are performed while the complexity of @xmath2 is still minimized .
satisfying the second sufficient condition from lemma [ l : move ] , these bigon moves preserve the fact that all components of @xmath193 are disks .
all remaining bigon moves follow arcs that intersect @xmath62 , meeting the first sufficient condition from lemma [ l : move ] .
therefore , these moves too preserve the fact that all components of @xmath193 are disks .
several proofs in [ s:3]-[s:4 ] will use the following lemma .
[ l : inessential ] if @xmath194 is an arc whose endpoints @xmath195 lie on distinct circles of @xmath95 , then these endpoints also lie on distinct circles of @xmath130 .
let @xmath194 be an arc whose endpoints @xmath195 lie the same circle @xmath12 of @xmath130 .
we claim that these endpoints must also lie on the same circle of @xmath95 .
since @xmath12 bounds a disk @xmath157 , there is an arc @xmath158 with the same endpoints as @xmath107 .
the circle @xmath159 lies on the boundary of some component of @xmath145 , a ball , and therefore bounds a disk @xmath160 whose interior is disjoint from @xmath2 .
the arc @xmath196 is parallel in the link exterior through @xmath114 to @xmath2 ; @xmath3-incompressibility implies that @xmath107 must also be parallel in @xmath62 to @xmath2 , and in particular that the endpoints of @xmath107 must lie on the same component of @xmath95 , as claimed .
the following special case is particularly noteworthy : [ l : inessentialcase ] the two arcs of @xmath95 traversing each over / underpass lie on distinct circles of @xmath130 .
this follows immediately from lemma [ l : inessential ] .
say that @xmath2 has a standard * tube * near a crossing ball @xmath31 if there are two arcs @xmath197 such that ( 1 ) for @xmath198 , there is an isotopy of @xmath199 through @xmath200 to @xmath201i.e . , for @xmath198
, @xmath202 is parallel through @xmath136 to @xmath31 , allowing the endpoints to slide along @xmath203and ( 2 ) the endpoints of @xmath204 , @xmath205 are also endpoints of the four arcs of @xmath134 that traverse the overpass and underpass at @xmath31i.e .
these endpoints are ( among ) the points of @xmath206 closest to @xmath31 in each of the four directions along @xmath62 , in the sense of the disk - bundle @xmath57
. features two arcs @xmath197 parallel through @xmath207 to @xmath31 whose endpoints are ( among ) the points of @xmath206 closest to @xmath31 in each direction along @xmath62 . in a minimal crossing link diagram
, such a tube gives a compressing disk @xmath114 for @xmath2 in @xmath9 with @xmath208 ( cf .
lemma [ l : sufficient ] ) , implying that @xmath209 . compressing @xmath2 along @xmath114 changes @xmath55 , @xmath2 as shown.,height=168 ] up to symmetry , this appears as in figure [ fi : smoothing]the arcs @xmath204 , @xmath205 must be in opposite quadrants relative to @xmath31 , not adjacent ones .
the reason for this is that the only other possibility ( cf .
figure [ fi : twoinnermosts ] , second from left ) contradicts the essentiality of @xmath2 in the link exterior , using lemma [ l : inessential ] , specifically lemma [ l : inessentialcase ] .
a crossing tube sets up a surgery move on @xmath2 , @xmath0 as follows .
each of the two arcs @xmath204 , @xmath205 associated with a crossing tube near @xmath31 has one endpoint on an edge - pair incident to the overpass at @xmath31 and the other on an edge - pair incident to the underpass at @xmath31 .
the two endpoints on edge - pairs incident to the overpass at @xmath31 can be joined by an arc @xmath210 ; likewise , the two endpoints on edge - pairs incident to the underpass at @xmath31 can be joined by an arc @xmath211 .
the circle @xmath212 bounds a disk @xmath108 whose interior lies in @xmath213 .
both @xmath214 , @xmath215 are parallel through disks @xmath216 to arcs @xmath217 each of which intersects @xmath0 in a single point .
thus , the disk @xmath218 satisfies @xmath219 with @xmath208 .
moreover , the arcs @xmath220 , @xmath221 can be isotoped so that the two points ( total ) in which they intersect @xmath0 are the endpoints of the vertical crossing arc in @xmath31 , one on the overpass and the other on the underpass .
the surgery move associated to the crossing tube near @xmath31 ( cf .
figure [ fi : smoothing ] ) consists of ( 1 ) cutting @xmath2 along @xmath222 , while cutting @xmath0 at the two points of @xmath223 , and ( 2 ) gluing in two parallel copies of the disk @xmath114 , while joining each pair of endpoints of @xmath0 on the boundary of the glued - in copy of the disk @xmath114 with an arc in that disk .
this surgery move is a compression of @xmath2 in @xmath9 unless @xmath222 bounds a disk in @xmath2 , in which case the surgery move yields a surface with two components , one of them a sphere .
the effect of the surgery move on @xmath0 is the same as one of two possible `` smoothings '' near @xmath31 , in the traditional sense from skein relations .
after the surgery move , the resulting link is again embedded in the resulting surface .
[ l : sufficient ] given a crossing tube in a minimal crossing diagram of a non - split link , the associated surgery move is a compression the boundary of the surgery disk @xmath114 does not bound a disk in @xmath2 .
in particular , a crossing tube in a reduced alternating link diagram contains a `` genuine '' compressing disk @xmath114 for @xmath2 in @xmath9 with @xmath208 .
this lemma will round off the proof of the main theorem in [ s:4 ] .
let @xmath31 be the crossing ball with the tube in question .
construct the disk @xmath114 as before , and suppose for contradiction that @xmath222 bounds a disk @xmath113 in @xmath2 . since @xmath224 and @xmath0 is non - split , @xmath225 consists of a single arc , call it @xmath182 .
from @xmath226 it follows that the two 2-spheres on the boundary of a thin regular neighborhood of @xmath227 intersect @xmath0 in a total of two points ; hence , one of these 2-spheres is disjoint from @xmath0 . since @xmath0 is non - split , this implies that the disks @xmath113 and @xmath114 are parallel through a ball @xmath186 in @xmath9 _ whose interior is disjoint from @xmath0_. the arc @xmath228 is parallel through the ball @xmath186 to any arc @xmath229 in @xmath114 that joins the overpass and underpass of @xmath0 at @xmath31 .
taking @xmath229 to be the vertical arc in @xmath31 that joins the overpass and underpass of @xmath0 , isotope the arc @xmath230 to @xmath229 , while fixing the rest of @xmath0 .
this isotopy eliminates all crossings incident to @xmath182 , including the one at @xmath31 , without creating any new ones .
we assumed this was impossible . in the broader
setting of [ s : broader ] , construct graphs @xmath231 ( they will be trees ) whose vertices correspond to the components of @xmath232 cut along @xmath2 and whose edges correspond to the components of @xmath130 , such that the edge corresponding to each circle @xmath233 joins the vertices corresponding to the two components of @xmath232 cut along @xmath2 whose boundaries contain @xmath12 .
the two `` sides of @xmath12 in @xmath232 '' refer to the two components of @xmath232 cut along @xmath12 and correspond to the two components of the graph @xmath234 obtained from @xmath231 by deleting the interior of the edge associated with @xmath12 .
define the * height * of @xmath12 on each side to be the _ maximum _ edge - length among all ( simple ) paths in the corresponding component of @xmath234 that start at the ( appropriate endpoint of the ) edge associated with @xmath12 .
thus , innermost circles of @xmath130 have height 0 ( to one side ) , circles which enclose ( to one side ) innermost circles and no others have height 1 , and so on .
figure [ fi : t34 ] shows an example a torus containing the knot @xmath29 in an almost - alternating diagram , with the associated graph @xmath235 .
[ p : height1exists ] either all circles of @xmath130 have height 0 ( to one side ) , or @xmath130 contains a circle with height 1 ( to one side ) .
let @xmath12 be a circle of @xmath130 with height @xmath236 ( to one side)if no such @xmath12 exists , there is nothing to prove .
consider a maximal edge - length path in @xmath237 that starts at the appropriate end of the edge associated with @xmath12 .
this path traverses @xmath238 edges , the @xmath239 of which corresponds to a circle of @xmath130 with height @xmath240 .
to review the setup and preliminary results , a link @xmath90 follows a link diagram @xmath40 with @xmath123 crossings ; @xmath55 is a regular neighborhood of @xmath0 , seen as ( the total space ) of a disk - bundle @xmath57 ; balls @xmath69 are the closures of the two components of @xmath70 , with @xmath91 and @xmath75 ; and a closed surface @xmath2 , which contains @xmath0 and is * essential * ( i.e. incompressible and @xmath3-incompressible in the link exterior , i.e. @xmath241 ) , has been isotoped such that : * the restriction @xmath93 is a bundle map ; * @xmath2 is transverse to @xmath73 , @xmath54 ; * @xmath96 ; and , subject to these conditions , the * complexity * of @xmath2 is * minimized * , specifically : * the numbers of components of @xmath97 and of @xmath98 have lexicographically been minimized . with this initial setup , [ s:24 ] showed that all components of @xmath142 , @xmath143 and @xmath242 are disks ; all components of @xmath134 , @xmath135 , and @xmath98 are arcs ; and all components of @xmath243 and @xmath244 are balls .
actually , the last two statements follow from the first one , using the essentiality of @xmath2 . in the somewhat more general setting where this first condition replaces the assumption of minimal complexity , [ s : bigon]-[s : height ] established such technical conveniences as bigon moves , crossing tubes , and height .
section [ s:3 ] delimits which local configurations are consistent with the initial setup , where @xmath2 is essential and its complexity is minimized .
most results address the arcs of @xmath134 , @xmath135 , and @xmath98 , and many extend to the case where @xmath2 is incompressible but @xmath3-compressible in the link exterior .
( one way to extend the proofs , roughly , is to isotope @xmath2 so as to push any @xmath3-compressing disks into @xmath55 , and then to slide these disks along the link away from the local area under consideration . )
several of the proofs require disrupting minimality by creating new components of @xmath98 , usually through a sequence of bigon moves .
there are several valid reasons to do this . in the proof of lemma [ l : edgecrossingloop ] ,
temporary _ increase in complexity enables the removal of a component of @xmath97 , lessening the complexity of @xmath2 , a contradiction .
one case in the proof of lemma [ l : twoinnermosts ] increases complexity in order to reveal a @xmath3-compressing disk , thus contradicting essentiality .
the other case in the proof of lemma [ l : twoinnermosts]and the proof of the crossing tube lemma in [ s:4]increase complexity in order to procure a crossing tube . in such cases ,
lemma [ l : move ] will confirm that the bigon moves , while disrupting minimality , preserve at least the fact that all components of @xmath142 , @xmath143 and @xmath242 are disks , and thus the more general setting of [ s : bigon]-[s : height ] . in particular , this will validate further bigon moves and applications of lemma [ l : inessential ] . assume throughout [ s:31 ] that @xmath32 is a diagram of a non - trivial , non - split link @xmath0 , and that @xmath0 is contained in a closed surface @xmath1 ( compact and connected without boundary ) .
establish all setup from [ s:21]-[s:23 ] .
assume in particular that @xmath2 is essential and its complexity @xmath99 has been minimized .
[ p : edgeloopb ] no arc of @xmath245 has both endpoints on the same component of @xmath82 .
[ p : crossingloopaa ] no arc of @xmath135 has both endpoints on the same component of @xmath246 . is parallel in @xmath62 to @xmath136 ( left ) , and no arc of @xmath135 is parallel in @xmath247 to @xmath136 ( right).,title="fig:",height=168 ] is parallel in @xmath62 to @xmath136 ( left ) , and no arc of @xmath135 is parallel in @xmath247 to @xmath136 ( right).,title="fig:",height=168 ] see figure [ fi : edgeloop ] . recall that @xmath248 by assumption .
any arc @xmath204 of @xmath134 with endpoints on the same component of @xmath82 must be parallel through a disk @xmath249 to an arc @xmath250 ; but then , isotoping @xmath2 near @xmath204 through @xmath108 past @xmath214 ( figure [ fi : edgeloop ] , left ) would reduce the complexity of @xmath2 , contrary to assumption .
likewise , any arc @xmath205 of @xmath135 with endpoints on the same component of @xmath246 must be parallel through a disk @xmath251 to an arc @xmath252 ; but then , isotoping @xmath2 near @xmath205 through @xmath113 past @xmath215 ( figure [ fi : edgeloop ] , right ) would reduce the complexity of @xmath2 , contrary to assumption . [
l : saddle ] every component of @xmath144 is a disk whose boundary consists of four arcs , alternately on @xmath253 and @xmath254 , none of which is parallel in @xmath255 to @xmath246 . , ( center ) intersecting @xmath2 in a single component , ( right ) intersecting @xmath2 in at least two components ( cf .
lemma [ l : saddle ] ) .
, height=168 ] that is , each component of @xmath144 looks like a saddle , as in figure [ fi : crossingsaddle ] ( center ) , and each crossing ball looks like one of the pictures in figure [ fi : crossingsaddle ] , depending on the number of components in which it intersects @xmath2 .
lemma [ l : saddle ] is an immediate consequence of propositions [ p : disks ] , [ p : arcs ] , and [ p : crossingloopaa ] .
[ p : edgeloope ] no arc of @xmath98 has both endpoints on the same component of @xmath82 .
[ p : crossingloope ] no arc of @xmath98 has both endpoints on the same crossing ball .
is parallel in @xmath136 to @xmath62 ( left ) or to @xmath247 ( right).,title="fig:",height=168 ] is parallel in @xmath136 to @xmath62 ( left ) or to @xmath247 ( right).,title="fig:",height=168 ] if both endpoints of some arc of @xmath98 were on the same component of @xmath82 then , applying proposition [ p : edgeloopb ] , an outermost such arc in @xmath136 would appear as left in figure [ fi : crossingloop ] , contradicting the assumed @xmath3-incompressibility of @xmath2 in the link exterior , e.g. by lemma [ l : inessential ] .
suppose instead that @xmath256 is an arc of @xmath98 with both endpoints on the same crossing ball @xmath31 , and assume that @xmath256 is outermost in @xmath136 , i.e. parallel through @xmath207 to @xmath31 , as in figure [ fi : crossingloop ] .
push @xmath2 near @xmath256 through @xmath136 past @xmath49 .
this attaches two saddle - shaped disks in the interior of @xmath31 , lessening the complexity of @xmath2 , contrary to assumption .
[ l : edgecrossingloop ] no arc @xmath257 has one endpoint on a crossing ball and the other on an incident edge of @xmath62 .
suppose that @xmath256 is an arc of @xmath98 with one endpoint on a crossing ball @xmath31 and the other on an incident edge of @xmath62 .
assume that @xmath256 is outermost in @xmath136 , i.e. parallel through @xmath207 to @xmath258 .
consider the circle of @xmath95 that contains an endpoint of @xmath256 . moving along this circle from that endpoint toward @xmath31
, there is at most one more point on @xmath136 , by proposition [ p : edgeloopb ] and the assumption that @xmath256 is outermost .
there are thus two cases up to symmetry ( cf .
figure [ fi : edgecrossingloop ] ) . of @xmath98 has one endpoint on a crossing ball and the other on an incident edge of @xmath62.,title="fig:",width=624 ] of @xmath98 has one endpoint on a crossing ball and the other on an incident edge of @xmath62.,title="fig:",width=624 ] in either case , begin with a sequence of ( up to ) three bigon moves , through the arcs labeled @xmath204 , @xmath205 , @xmath259 in figure [ fi : edgecrossingloop ] , in that order .
any of the arcs @xmath202 can be parallel in @xmath150 to @xmath2 ; in this case , omit the bigon move along @xmath202 . in all cases ,
this sequence of bigon moves fits the hypotheses of lemma [ l : movecase ] and thus preserves the fact that all components of @xmath193 are disks .
now an arc @xmath260 is parallel through a disk @xmath261 with @xmath262 to a second arc @xmath263 with @xmath264 .
this arc @xmath265 is parallel to an arc @xmath266 through a disk @xmath160 with @xmath267 .
further , @xmath268 is parallel to @xmath109 through a disk @xmath269 , which contains an entire disk of @xmath144 . finally , the 2-sphere @xmath270 bounds a ball @xmath186 in the link exterior .
isotope @xmath271 through @xmath272 to @xmath273 , while fixing @xmath274 .
this removes the disk of @xmath275 and thus a component of @xmath276 .
since bigon moves always fix @xmath277 , this contradicts the initial assumption that the complexity of @xmath2 was minimized .
maintain all setup from [ s:31 ] , with the additional assumption that @xmath32 is alternating .
that is , assume throughout [ s:32 ] that @xmath32 is a non - trivial , connected , alternating diagram of a link @xmath0 , and that @xmath0 is contained in a closed essential surface @xmath1 ( compact and connected without boundary ) whose complexity @xmath99 has been minimized .
when @xmath32 is alternating ( cf .
lemma [ l : edgetypes]).,width=624 ] [ l : edgetypes ] every edge - pair of @xmath62 appears as in figure [ fi : edgetypes ] .
this follows immediately from propositions [ p : arcs ] and [ p : edgeloopb ] and the alternatingness of @xmath32 .
[ p : innermostcrossingball ] if @xmath12 is an innermost circle of @xmath130 , then @xmath278 .
is disjoint from @xmath247 ( cf .
proposition [ p : innermostcrossingball]).,width=624 ] let @xmath279 be an innermost circle of @xmath280 ( wlog ) , and suppose @xmath281 .
then the disk @xmath282 with @xmath283 intersects @xmath247 ; let @xmath113 be a component of @xmath284 .
there are now three cases , using lemma [ l : saddle ] ( cf .
figure [ fi : innermostcrossingball ] ) : in one case ( left ) , @xmath285 contains two arcs of @xmath286 .
otherwise @xmath285 contains an arc on @xmath62 , along either the underpass ( center ) or the overpass ( right ) at @xmath31 . if @xmath285 contains more than one arc of @xmath287 ( left in figure [ fi : innermostcrossingball ] ) , then a bigon move yields an arc of @xmath98 with both endpoints on @xmath31 .
this contradicts the minimality of @xmath97 ( recall figure [ fi : crossingloop ] , right , and the proof of proposition [ p : crossingloope ] ) . in both remaining cases ( center and right ) , a bigon move yields an arc of @xmath98 with one endpoint on @xmath31 and the other on an incident edge of @xmath62 .
crucial to the initial bigon moves in figure [ fi : innermostcrossingball ] , center and right , is that every edge of @xmath62 contains endpoints of @xmath98 , by alternatingness .
next , perform the sequence of bigon moves from the top of figure [ fi : edgecrossingloop ] ( omitting any trivial ones as usual ) , to set up the final isotopy move from that sequence .
this final move is valid , since the preceding sequence of bigon moves meets the conditions of lemma [ l : movecase ] , and it removes a component of @xmath276 .
since bigon moves always fix @xmath277 , this contradicts the initial assumption that the complexity of @xmath2 was minimized .
[ p : innermostpass ] if @xmath12 is an innermost circle of @xmath130 , so that @xmath12 bounds a disk @xmath288 , then at least one component of @xmath289 traverses an over / underpass .
lemma [ l : edgetypes ] implies that the only arcs of @xmath95 with endpoints on the same component of @xmath290 look like the arcs of this type in figure [ fi : edgetypes ] , up to reflection there is one far left in figure [ fi : edgetypes ] , one second from right , and two far right . if such an arc lies on an innermost circle @xmath12 of @xmath280 ( wlog ) , then , since @xmath12 is innermost and @xmath278 by proposition [ p : innermostcrossingball ] , @xmath12 must traverse the overpass at the crossing where the edge of @xmath62 containing this arc of @xmath12 meets an underpass .
this is evident in figure [ fi : edgetypes ] , using lemma [ l : inessential ] .
[ l : twoinnermosts ] if both circles of @xmath130 traversing a given over / underpass have height 0 ( to one side ) , i.e. are innermost , then @xmath2 can be isotoped to have a standard tube near that crossing . traversing a given over / underpass , say at @xmath31 ,
have height zero , then @xmath2 can be isotoped to have a standard tube near @xmath31.,title="fig:",width=288 ] traversing a given over / underpass , say at @xmath31 , have height zero , then @xmath2 can be isotoped to have a standard tube near @xmath31.,title="fig:",width=288 ] let @xmath31 be the ball at the crossing in question .
recall that the two arcs traversing the overpass ( wlog ) at @xmath31 lie on distinct circles of @xmath280 , by lemma [ l : inessentialcase ] .
consider the two edge - pairs of @xmath62 that are incident to the underpass at @xmath31 . because @xmath32 is alternating
, each edge - pair contains endpoints of @xmath98 .
consider the endpoint on each edge - pair that is nearest to @xmath31 in @xmath62recall that @xmath95 is ( the image of ) a section of the disk - bundle @xmath291 .
up to symmetry , there are two cases , depending on which sides of @xmath32 these two points lie on , relative to each other ( cf . figure [ fi : twoinnermosts ] ) . if these two points lie in adjacent quadrants near @xmath31 ( left in figure [ fi : twoinnermosts ] ) , perform two bigon moves ( unless the associated arc is parallel in @xmath150 to @xmath2 ) .
since each ( possible ) bigon move follows an arc with an endpoint on @xmath62 , lemma [ l : move ] implies that these moves preserve the fact that @xmath292 cuts @xmath2 into disks .
moreover , they produce a diagram in which an arc @xmath293 has endpoints on the same circle of @xmath294 but on distinct circles of @xmath95 , which lemma [ l : inessential ] states is impossible .
therefore , these two points must lie in opposite quadrants @xmath295 near @xmath31 ( right in figure [ fi : twoinnermosts ] ) . in this case , a pair of bigon moves ( to be omitted if trivial ) immediately fashions a standard tube near @xmath31 . lemma
[ l : movecase ] implies that these moves preserve the fact that @xmath292 cuts @xmath2 into disks .
given a non - trivial , connected , reduced alternating diagram of a link @xmath0 and a closed , essential surface @xmath36 , there exists an isotopy after which @xmath2 has a standard tube near some crossing . as in [ s:2 ]
, let @xmath90 follow a reduced alternating diagram , with @xmath55 a closed regular neighborhood of @xmath0 seen as ( the total space of ) a disk - bundle @xmath57 , @xmath69 the components of @xmath296 , and @xmath91 .
let @xmath2 be a closed , essential surface containing @xmath0 .
fixing @xmath92 , @xmath33 , and @xmath38 , isotope @xmath2subject to the requirements that @xmath297 ; @xmath93 is a bundle map ; and @xmath96so as to minimize lexicographically the numbers of components of @xmath97 and @xmath98 .
consider @xmath280 .
if all circles have height 0 , apply lemma [ l : twoinnermosts ] at any overpass , done . otherwise , by proposition [ p : height1exists ] , there exists a circle @xmath298 of @xmath280 with height 1 .
let @xmath279 be any ( innermost ) circle enclosed by @xmath298 .
apply proposition [ p : innermostpass ] to consider an overpass which @xmath279 traverses .
let @xmath12 denote the other circle of @xmath280 traversing this overpass .
note that @xmath299 by lemma [ l : inessential ] , specifically lemma [ l : inessentialcase ] . if @xmath12 has height 0 , then lemma [ l : twoinnermosts ] completes the proof . otherwise , @xmath12 must equal @xmath298 .
see figure [ fi : finalmove ] . next , consider the circle @xmath171 of @xmath280 from figure [ fi : finalmove ] , which must exist and be distinct from @xmath279 , due to lemmas [ l : edgetypes ] ,
[ l : inessential ] , and the assumption that @xmath279 has height 0 . if @xmath171 also has height 0 , then lemma [ l : inessential ] implies that the edge of @xmath62 in question must appear as in figure [ fi : edgetypes ] , second from right
; thus , @xmath279 and @xmath171 must traverse a common overpass , completing the proof , using lemma [ l : twoinnermosts ] .
otherwise , @xmath300 .
this allows the sequence of isotopy moves shown in figure [ fi : finalmove ] , yielding the desired crossing tube .
( again , omit either isotopy move if the associated arc is parallel in @xmath150 to @xmath2 ; and lemma [ l : movecase ] applies since each bigon move is along an arc with an endpoint on @xmath62 . ) every non - split , non - trivial alternating link @xmath0 has representativity @xmath26 . beginning with a reduced alternating diagram of @xmath0 ,
apply the crossing tube lemma to obtain a standard tube at some crossing . then apply lemma [ l : sufficient ] to conclude that the crossing tube contains a disk @xmath114 with @xmath301 , such that @xmath222 intersects @xmath0 in two points and does not bound a disk in @xmath2 .
containing the knot @xmath29 in an almost - alternating diagram , with the graph @xmath235 .
there are two circles of @xmath280 with height 2 , two with height 1 , and six with height 0.,height=336 ] figure [ fi : t34 ] shows a torus @xmath2 containing the knot @xmath302 , with @xmath0 in an almost - alternating diagram .
the torus is difficult to visualize directly from this diagram , but its homeomorphism type is straightforward to verify by computing euler characteristic . as @xmath303 , this example illustrates how the arguments leading up to the crossing tube lemma break down without alternatingness .
the configuration satisfies the conclusions of all propositions and lemmas from [ s:31 ] , but escapes those from [ s:32 ] ( regarding edge types and innermost possibilities ) .
if an almost - alternating link @xmath0 has representativity @xmath304 , then there can be no crossing tube in any diagram with minimal crossing number .
thus , after applying the setup and results from 2 , 3.1 , all non - innermost circles of @xmath280 and @xmath294 must ( properly enclose or ) be incident to the `` de - alternator . '' otherwise , a local application of the proof of the crossing tube lemma gives @xmath305 .
likewise , at least one of the two circles traversing each over / underpass must be incident to ( or properly enclose ) the de - alternator .
perhaps it is feasible to list explicitly which almost alternating links have representativity @xmath30 .
99 c. adams , c. , t. kindred , _ a classification of spanning surfaces for alternating links _
topology 13 ( 2013 ) , no . 5 , 2967 - 3007 .
arxiv:1205.5520 r. blair , m. campisi , s. taylor , m. tomova , private communication .
d. gabai , _ genera of the alternating links _ , duke math j. vol 53 ( 1986 ) ,
3 , 677 - 681 .
m. gromov , _ filling riemannian manifolds _ , j. differential geom . 18 ( 1983 ) , no . 1 , 1 - 147 .
w. menasco , _ closed incompressible surfaces in alternating knot and link complements _ , topology 23 ( 1984 ) , no
. 1 , 37 - 44 .
m. ozawa , _ the representativity of pretzel knots _ , j. knot theory ramifications 21 ( 2012 ) , no . 2 , 1250016 , 13 pp .
arxiv:0911.2979 m. ozawa , _ non - triviality of generalized alternating knots _ , j. knot theory ramifications 15 ( 2006 ) , no . 3 , 351 - 360
arxiv : math/0504012 m. ozawa , _ bridge position and the representativity of spatial graphs _ , topology appl .
159 ( 2012 ) , no . 4 , 936 - 947 .
arxiv:0909.1162 j. pardon , _ on the distortion of knots on embedded surfaces _ ,
ann . of math .
( 2 ) 174 ( 2011 ) , no . 1 , 637 - 646 . | we prove that if @xmath0 is a non - trivial alternating link embedded ( without crossings ) in a closed surface @xmath1 , then @xmath2 has a compressing disk whose boundary intersects @xmath0 in no more than two points .
moreover , whenever the surface is incompressible and @xmath3-incompressible in the link exterior , it can be isotoped to have a standard tube at some crossing of any reduced alternating diagram . |
in nature , complex and chaotic flow universally appears in various systems from small to large scale , for instance , in watercourses , rivers , oceans , and the atmosphere .
such a flow , called turbulence , has attracted considerable attention in various fields because understanding it leads to significant development of fundamental and applied sciences such as mathematics , physics , and technology .
though turbulence has been studied for a long time , however , it is considered to be one of the unresolved problems and becomes an important issue in modern physics .
originally , turbulence has been studied in classical fluids , where it is important to investigate statistical quantities such as the probability density function of the velocity field @xcite .
one of the most significant quantities is the kinetic energy spectrum corresponding to a two - point correlation function of the velocity field , which is known to exhibit the famous kolmogorov @xmath2 power law .
this kolmogorov law appears as a result of the constant - flux energy cascade in wave number space , and the observation of this power law is very important in the study of turbulence . in quantum fluids , turbulence also appears and has been investigated for many years @xcite .
this turbulence is called quantum turbulence ( qt ) , in contrast to classical turbulence ( ct ) , and was originally developed in the field of superfluid helium .
qt offers rich phenomena such as thermal counterflow @xcite , richardson cascade @xcite , kelvin wave cascade @xcite , the bottleneck effect @xcite , velocity statistics @xcite , and the reconnection of quantized vortices @xcite , leading to a deep understanding of turbulence , and it has been actively studied until today .
recently , atomic bose - einstein condensates ( becs ) have been providing a novel stage for turbulence studies in quantum fluids @xcite .
this system has three advantages for investigating turbulence in quantum fluids .
the first advantage is the experimental visualization of the density and vortex distribution @xcite , which gives us a large amount of information about turbulence .
such an observation is difficult in superfluid helium , though the vortices have been visualized recently by use of tracer particles @xcite .
the second advantage is the high controllability of the system by optical techniques @xcite , which enable us to change the shape of the quantum fluid and the interaction between particles .
this can access new problems : the geometrical dependence of turbulence such as two - dimensional qt .
this kind of study is being actively investigated .
the third advantage is the realization of multicomponent becs such as binary @xcite and spinor becs @xcite .
the hydrodynamics in this kind of quantum fluid will give us novel physics , and investigations in this area have recently begun @xcite .
thus , turbulence in the atomic becs can address exotic phenomena not found in superfluid helium . at present
, there is some research on turbulence in atomic becs , and most theoretical and numerical studies consider turbulence with many quantized vortices . in three- and two - dimensional systems , when many quantized vortices
are nucleated , the kolmogorov @xmath2 power law is numerically confirmed in incompressible kinetic energy @xcite .
recently , two - dimensional qt has been actively investigated , with discussion of features characteristic of two - dimensional systems such as inverse energy cascades @xcite , vortex clustering @xcite , negative temperature @xcite , and decay of vortex number @xcite . in experiments ,
some groups have succeeded in generating turbulence in atomic becs @xcite , where it has been possible to obtain turbulence with many quantized vortices .
all these experimental studies focus on quantized vortices , investigating the anomalous expansion @xcite , vortex dynamics @xcite , annihilation of vortices @xcite , and so on . however , the kinetic - energy spectrum corresponding to a two - point correlation function of the velocity field has not yet been observed , and we can not confirm the kolmogorov @xmath2 power law .
this means that kolmogorov turbulence has not been observed .
let us elucidate the meaning of kolmogorov turbulence .
in the usual hydrodynamic turbulence ( ht ) , the vortices give important structures , whose dynamics generates the @xmath2 power law in the kinetic - energy spectrum due to a constant - flux cascade for the kinetic energy .
however , in other systems , various power laws appear as a result of the constant - flux cascade for some quantity such as the wave energy , wave action , and so on . in these cases ,
the common physics for the appearance of the power law is the constant - flux cascade . in this paper , we call such turbulence with the constant - flux cascade kolmogorov turbulence . in the turbulence study
, it would be very important to observe kolmogorov turbulence ; however , observing the superfluid velocity directly in turbulence of atomic becs is difficult , so we can not confirm whether kolmogorov turbulence appears as of now . to shed light on the above problem , we focus on weak wave turbulence ( wt ) with a strong condensate in atomic becs because the density profile of the bec is observable and thus plays an important role in observing kolmogorov turbulence . in this turbulence ,
bogoliubov waves are significant , so we call it bogoliubov - wave turbulence in this paper
. originally , in classical fluids , wt , which is turbulent flow dominated by waves , has been studied @xcite .
it is much different from ht because vortex structures are regarded as important in ht @xcite .
there are weak and strong kinds of wt , depending on whether the nonlinearity is weak or strong . in weak wt ,
a constant - flux cascade of wave energy or wave action generates power - law behaviors in the spectrum , and kolmogorov turbulence is known to occur in various wave systems such as those involving fluid surfaces ( gravity and capillary waves ) @xcite , magnetic substances ( spin waves ) @xcite , and elastic media ( acoustic waves ) @xcite .
we expect that it is possible to confirm kolmogorov turbulence in atomic becs by the observation of the density profile in bogoliubov - wave turbulence . in this paper , by applying weak wt theory @xcite to the gross - pitaevskii ( gp ) equation , we theoretically and numerically study the power - law behaviors for three spectra of a macroscopic wave function , density distribution , and bogoliubov wave distribution , discussing the experimental possibility for observing kolmogorov turbulence by means of the density spectrum . there are some previous works on bogoliubov - wave turbulence in three - dimensional systems @xcite . in previous studies
@xcite the @xmath0 power law in the bogoliubov - wave energy spectrum was analytically derived , and the equation of the fluctuation was derived with the assumption that the condensate function has a constant amplitude with a phase rotation induced by the chemical potential .
subsequent work @xcite has suggested the @xmath0 power law in the spectrum for the macroscopic wave function and a result consistent with this power law was numerically obtained .
we reconsider this bogoliubov - wave turbulence , analytically deriving the @xmath1 power law in the spectrum for the macroscopic wave function and numerically confirming this power law .
we find that the condensate dynamics induced by the fluctuation , which is neglected in the previous studies , is important for bogoliubov - wave turbulence .
furthermore , we focus on the spectrum for the density distribution , obtaining the @xmath0 power law , and discuss the experimental possibility of this power law .
the article is organized as follows .
section ii describes the gp equation and the spectra for some quantities . in sec .
iii , we apply weak wt theory to the gp equation , deriving the power laws for the spectra . in sec .
iv , we show our numerical result for this bogoliubov - wave turbulence .
section v discusses comparison between the previous results and ours , the observation of kolmogorov turbulence , and an inconsistency of our analytical calculation .
finally , we summarize our study in sec . vi .
we address a one - component bec in a uniform system at zero temperature .
this system is well described by the macroscopic wave function @xmath3 obeying the gp equation @xcite given by @xmath4 with particle mass @xmath5 and interaction coefficient @xmath6 . in this paper , we focus on spectra for the macroscopic wave function @xmath3 , the density distribution @xmath7 , and the bogoliubov - wave distribution @xmath8 .
the details of the bogoliubov - wave distribution are defined in sec .
iii b. the spectrum for the wave function is defined by @xmath9 with a resolution of @xmath10 in wave - number space and a system size @xmath11 .
the brackets indicate an ensemble average .
the function @xmath12 is the fourier component of the macroscopic wave function calculated by @xmath13 $ ] with @xmath14 = \int \cdot \hspace{1 mm } e^{-i\bm{k } \cdot \bm{r } } dv / l^d$ ] and spatial dimension @xmath15 . in the same way
, we can define the spectra for the density and bogoliubov - wave distributions as @xmath16 @xmath17 with the fourier components of the density distribution @xmath18 $ ] and the bogoliubov - wave distribution @xmath19 .
the physical meaning of these spectra is that @xmath20 and @xmath21 are correlation functions for the macroscopic wave function and density profile , and @xmath22 is the bogoliubov - wave distribution in the wave - number space .
we study weak wt in a uniform bec with the strong condensate , deriving the power exponents of @xmath23 , @xmath24 and @xmath25 . for this purpose , we apply weak wt theory to the gp equation ( [ gp ] ) . in this application
, there is an important points , which is the nonlinear dynamics for the condensate .
we take the squared terms of the fluctuation for the macroscopic wave function , treating this nonlinear dynamics . in this case , we can also use weak wt theory , obtaining these power exponents .
we now derive the equations of the condensate and the fluctuation .
first , we define the condensate @xmath26 and the fluctuation @xmath27 as @xmath28 where @xmath29 is defined by @xmath30 from eqs .
( [ fluctuation ] ) and ( [ condensate ] ) , we can derive @xmath31 equations ( [ condensate ] ) and ( [ fluctuation1 ] ) mean that the condensate and the fluctuation are the @xmath32 and @xmath33 fourier components of @xmath12 , respectively . in this paper
, we consider a weak fluctuation from a strong condensate , assuming the weak nonlinear condition @xmath34 .
we substitute eq .
( [ fluctuation ] ) into the gp equation ( [ gp ] ) , deriving equations for @xmath29 and @xmath35 $ ] within second order of the fluctuation : @xmath36 , \label{gp_condensate}\end{aligned}\ ] ] @xmath37 , \label{gp_fluctuation1}\end{aligned}\ ] ] where @xmath38 is the condensate density @xmath39 and @xmath40 is the kronecker delta . substituting eq .
( [ gp_condensate ] ) into eq .
( [ gp_fluctuation1 ] ) , we obtain the following equation for the fluctuation : @xmath41 . \label{gp_fluctuation2}\end{aligned}\ ] ] therefore , we obtain eqs .
( [ gp_condensate ] ) and ( [ gp_fluctuation2 ] ) for the condensate and the fluctuation dynamics within the weak nonlinear condition . for application of weak wt theory , we rewrite eq .
( [ gp_fluctuation2 ] ) into the canonical form given by @xmath42 @xmath43 @xmath44 \label{hamiltonian2 } , \end{aligned}\ ] ] @xmath45 here , note that the condensate function @xmath29 has a time dependence . in previous works @xcite ,
the condensate function @xmath29 was assumed to be @xmath46 with condensate density @xmath47 , the @xmath32 particle number @xmath48 at the initial state , and the chemical potential @xmath49 .
however , we keep the second - order fluctuation terms in eq .
( [ gp_condensate ] ) , so that the time dependence of @xmath29 is rather complicated .
this term is very important for calculating the bogoliubov - wave distribution , which is discussed in sec .
iv b. to diagonalize the one - body hamiltonian @xmath50 , we use the bogoliubov transformation @xcite defined by @xmath51 @xmath52 @xmath53 where @xmath19 is the canonical variable for bogoliubov waves ( bogoliubov - wave distribution ) and @xmath54 is the dispersion relation for this wave defined by @xmath55 with @xmath56 . applying this transformation to the hamiltonian of eqs .
( [ hamiltonian2 ] ) and ( [ hamiltonian3 ] ) leads to @xmath57 @xmath58 where @xmath59 and @xmath60 are the interaction functions for bogoliubov waves .
these functions are given by @xmath61 @xmath62 where we use simple expressions such as @xmath63 and @xmath64 for @xmath65 and @xmath66 .
we must notice that the bogoliubov coefficients @xmath67 and @xmath68 have time dependence through the condensate density @xmath69 .
thus , the transformation of eq .
( [ bogo_trans1 ] ) is not the canonical transformation .
however , as shown in the following , the time dependence of these coefficients are found to give only small corrections in the weak nonlinear condition , so that the transformation of eq .
( [ bogo_trans1 ] ) approximately becomes the canonical transformation . to indicate this
, we derive the time development equation for @xmath38 from eq .
( [ gp_condensate ] ) , which is given by @xmath70 \label{density_equ}.\end{aligned}\ ] ] then , the time derivative of the bogoliubov coefficient @xmath67 becomes @xmath71 which means that the value of this derivative is the second order of the fluctuation .
the same thing is true in the time derivative of @xmath68 . as a result , in the weak nonlinear condition , the time derivative of eq .
( [ bogo_trans1 ] ) is @xmath72 since the third order of the fluctuation is neglected in eq .
( [ gp_fluctuation2 ] ) .
thus , the bogoliubov transformation of eq .
( [ bogo_trans1 ] ) approximately becomes the canonical transformation .
based on the above calculation , the time development equation for @xmath19 becomes @xmath73 this equation can describe weak nonlinear dynamics for bogoliubov waves . in the following subsection ,
we apply weak wt theory to eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) , discussing the behavior of the spectra for the wave function , density distribution , and bogoliubov - wave distribution . in weak wt theory , the kinetic equation for the spectrum of the canonical variable @xmath19
can be derived @xcite .
the spectrum is defined by @xmath74 applying weak wt theory to eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) , we obtain the following kinetic equation : @xmath75 @xmath76 with the dirac delta function @xmath77 . in this application , the contributions from @xmath78 and @xmath79 terms in eq .
( [ hamiltonian5 ] ) are found to be smaller than those of other terms @xcite because these terms accompany @xmath80 . in bogoliubov waves ,
the argument in this dirac delta function is never satisfied , so that the interaction @xmath81 does not appear in eqs .
( [ kin_equ1 ] ) and ( [ kin_equ2 ] ) .
the details of this calculation are described in @xcite , where a nonlinear canonical transformation causes the interaction @xmath81 to vanish .
the transformation does not change the expression for @xmath82 , but it does change the expression for @xmath19 . however , this change gives only a small contribution to @xmath83 .
note that the application of weak wt theory to eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) is slightly different from usual cases where the dispersion relation @xmath84 and the interaction @xmath59 do not have time dependence . in our case ,
the condensate density @xmath69 depends on time , which makes @xmath84 and @xmath59 dependent on time .
thus , we can not use the conventional weak wt theory in @xcite , but we can derive the kinetic equation ( [ kin_equ1 ] ) .
the detail of this derivation is described in appendix a. we are interested in the large - scale dynamics of bogoliubov waves because , in experiments , it may be difficult to observe small - scale dynamics . here , to elucidate the meaning of large scale ,
let us note the dispersion relation for bogoliubov waves . from eq .
( [ bog_dis ] ) , it follows that the @xmath85 dependence of the dispersion relation drastically changes depending on whether or not the amplitude of the wave number is larger than @xmath86 . in the wave - number region smaller than @xmath87 , the dispersion relation becomes phononlike , which is expressed by @xmath88 with sound velocity @xmath89 , while in the region larger than @xmath87 it behaves like that of a free particle . from this property of the dispersion relation , in this paper , the large ( small ) scale means the wave - number region is smaller ( larger ) than @xmath87 . finally , we must comment on the importance of the dimension of the kinetic equation in the low - wave - number region where @xmath54 has linear dispersion . in this region ,
bogoliubov - wave turbulence is similar to acoustic turbulence , for which , in a two - dimensional system , the validity of the kinetic equation is controversial @xcite .
thus , in the following , we address three - dimensional systems . in the next section , to investigate the behavior of the spectra in the low - wave - number region , we calculate an expression of the interaction @xmath59 for the low - wave - number limit .
we now derive the expression for @xmath82 in the low - wave - number region .
first , we assume that this interaction is local in the turbulent state , which means that the dominant contribution to the interaction comes from wave numbers of the same order @xcite .
this locality is also discussed in appendix d. thus , the wave - number region @xmath90 ( @xmath91 ) is considered in the following . in the low - wave - number region , the bogoliubov coefficients @xmath67 and @xmath68
are approximated as @xmath92 , \label{bogo_coe1}\end{aligned}\ ] ] @xmath93 .
\label{bogo_coe2}\end{aligned}\ ] ] we substitute eqs .
( [ bogo_coe1 ] ) and ( [ bogo_coe2 ] ) into the interaction @xmath59 of eq .
( [ interaction1 ] ) , obtaining the following expression : @xmath94 .
\label{interaction3}\end{aligned}\ ] ] as shown by the two delta functions of eq .
( [ kin_equ2 ] ) , bogoliubov waves must satisfy both momentum- and energy - conservation laws , which are expressed by @xmath95 and @xmath96 .
these conservations lead to the rather simple form of eq .
( [ interaction3 ] ) . in the low - wave - number region ,
the dispersion relation becomes @xmath97 .
\label{disper}\end{aligned}\ ] ] as a result , the effective interaction @xmath98 for bogoliubov waves becomes @xmath99 the derivation of this interaction is described in appendix b. this interaction satisfies the scaling law @xmath100 , which is directly related to the power exponent of the spectra .
in previous studies different canonical variables were used to derive the interaction function , which is different from eq .
( [ interaction4 ] ) , but the scaling law is the same @xcite . in a stationary state , we can derive the power exponents of the spectra by using the kinetic equation ( [ kin_equ1 ] ) with the interaction of eq .
( [ interaction4 ] ) .
our derivation makes use of the fact that a bogoliubov - wave energy flux is constant in the wave - number space in the weak wt .
we integrate the kinetic equation ( [ kin_equ1 ] ) with the effective interaction ( [ interaction4 ] ) and the dispersion relation ( [ disper ] ) , obtaining @xmath101 @xmath102 dk_1 , \label{bogkin2}\end{aligned}\ ] ] @xmath103 @xmath104 with the heaviside step function @xmath105 . here
we assume the isotropy of @xmath106 , which is reasonable in isotropic turbulence .
in appendix c , we show the derivation of eqs .
( [ bogkin1 ] ) - ( [ bogkin4 ] ) . from eq .
( [ bogkin1 ] ) , we can derive the continuity equation of the bogoliubov - wave energy spectrum given by @xmath107 where @xmath108 and @xmath109 are the bogoliubov - wave energy spectrum and the energy flux defined by @xmath110 @xmath111 dk_1 .
\label{flux2}\end{aligned}\ ] ] in the statistically stationary state , the time derivatives of @xmath108 in eq .
( [ flux1 ] ) is zero , which means that the energy flux @xmath109 is independent of the wave number .
also , the time derivative of @xmath112 is zero because the condensate density @xmath69 is stationary in this state . assuming the power law @xmath113 and applying the transformation @xmath114 to eq .
( [ flux2 ] ) , we obtain @xmath115 which exhibits that the energy flux is constant if the power exponent @xmath5 is @xmath116 .
thus , we derive the power law in the bogoliubov - wave turbulence : @xmath117 this power law is found to be the local , so that the locality assumption in sec .
iv d is justified , which is described in appendix d. therefore , the spectrum integrated over the solid angle yields @xmath118 we can derive the power exponent for @xmath20 by using eq .
( [ b_sol2 ] ) . in the low - wave - number region
, the relation between @xmath119 and @xmath19 becomes @xmath120 which is obtained from eq .
( [ bogo_trans1 ] ) .
then , we assume isotropy for @xmath19 and use the approximation to derive eq .
( [ kin_equ1 ] ) , obtaining @xmath121 as a result , in the low - wave - number region , the spectrum for the macroscopic wave function exhibits the power law behavior given by @xmath122 in this derivation , we use the fourier component @xmath123 = \psi_0 ( \delta ( \bm{k } ) + \bar{\phi}(\bm{k}))$ ] . for the density spectrum ,
the following expression for @xmath3 is useful to derive the power exponent of @xmath21 : @xmath124 where @xmath29 is written as @xmath125 , and @xmath126 and @xmath127 are the density and phase fluctuations , respectively , around it .
then , the fluctuations of the wave function and the density are obtained from @xmath128 @xmath129 from which we can express the fourier component of @xmath126 with the canonical variable @xmath8 . performing a similar calculation for eqs .
( [ b_w_rel1])([b_w_rel3 ] ) , we can derive the power law in the spectrum for the density distribution : @xmath130 we consider that this power law is a candidate for an experimentally observable quantity to confirm the presence of kolmogorov turbulence .
as shown in above , the power exponents of @xmath20 and @xmath21 are related to @xmath22 through the behavior of bogoliubov coefficients of eqs .
( [ bogo_trans2 ] ) and ( [ bogo_trans3 ] ) in the low wave number region .
thus , the difference of the power exponents reflects the property of the collective mode . in the next section , by numerically calculating the gp equation
, we discuss these power laws .
we now present our numerical results for bogoliubov - wave turbulence with the gp equation .
one of the main results is in good agreement with the power exponents of eqs .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) derived by using weak wt theory .
[ b ] at @xmath131 ( a ) 0 , ( b ) 600 , ( c ) 1200 , and ( d ) 2500 .
the smaller structures are nucleated in turn as time passes , providing evidence of the direct energy cascade .
, width=321 ] [ t ] and @xmath132 for real and imaginary parts of the fluctuation @xmath27 .
these graphs exhibit a system satisfying the weak nonlinear condition .
, width=340 ] [ t ] and phase @xmath133 of the condensate @xmath29 .
the macroscopic wave function is found not to obey @xmath134 .
particularly , the deviation of the phase is large .
, width=340 ] our numerical calculation treats a three - dimensional bec without a trapping potential .
time and length are normalized by @xmath135 and @xmath136 , and the numerical system size @xmath137 is @xmath138 with spatial resolution @xmath139 . because of this resolution , the quantized vortex dynamics such as nucleation , reconnection , etc . can not be described correctly since its core size is the order of @xmath140 .
however , we address the turbulence without the vortices , so that it is sufficient to use this resolution parameter . in this situation , we numerically solve the gp equation by using the pseudo - spectral method .
the time propagation is performed by using a fourth - order runge - kutta method with time resolution @xmath141 .
we use an initial state in which energy is injected in the large scale to confirm the direct energy cascade solutions corresponding to eq .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) .
the initial state is prepared by the random numbers as follows : @xmath142 @xmath143(\bm{k } ) = 3000(r_1 + i r_2)k^4\xi^4 { \rm exp}(-k^2 \xi ^2/0.0016 ) , \end{aligned}\ ] ] where @xmath144 and @xmath145 are the random numbers in the range @xmath146 .
figure 1(a ) shows the density distribution in the initial state and demonstrates that the energy is injected at the large scale . in our calculation , dissipation
is phenomenologically included by replacing @xmath147 with @xmath148 in the fourier transformed eq .
( [ gp ] ) @xcite . here
, @xmath149 is @xmath150 , and @xmath151 is the strength of the dissipation . in our calculation , we use @xmath152 .
we comment on the dissipation region . in this paper , we focus on the low - wave - number region , adopting the dissipation working in the region larger than @xmath149 .
there is arbitrariness in the choice of the expression for the dissipation .
however , in a previous paper @xcite , the dissipation was found to work in the high - wave - number region at low temperature , so that we use this expression for the dissipation . in summary ,
we prepare the unstable initial state for the confirmation of the direct cascade , numerically calculating the gp equation with the phenomenological dissipation . in this calculation
, we do not force the system , so that this is decaying turbulence .
we numerically confirm that our turbulence satisfies the weak nonlinear condition @xmath153 .
figure 2 shows the time dependence of the spatially averaged absolute values @xmath154 and @xmath132 for real and imaginary parts of the fluctuation @xmath27 , from which the order of the fluctuation @xmath155 is found to be about @xmath156 . figure 3 shows the time development of the amplitude @xmath38 and phase @xmath133 of the condensate @xmath29 ; we find that the condensate does not obey @xmath134 .
this means that the fluctuation term in eq .
( [ gp_condensate ] ) affects the dynamics of the condensate .
this effect is very important for calculating the bogoliubov - wave distribution @xmath19 , which is discussed in sec .
iv c. [ t ] , the macroscopic wave function @xmath20 , and the density distribution @xmath21 .
the spectra are averaged over six calculations with different initial noise .
the direct energy cascade occurs , and the power exponents of eq .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) show good agreement with the numerical results .
, width=306 ] we next numerically calculate the spectra @xmath22 , @xmath20 , and @xmath21 , finding that the numerical results show good agreement with our analytical results of eqs .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) .
figure 4 shows the time development of the three spectra , demonstrating the direct energy cascade from small to large wave numbers .
this cascade is also confirmed by the density distribution as shown in figs .
[ fig1](a)[fig1](d ) , in which small - scale structure is seen to grow . in each spectrum , as time passes , the power - law behaviors corresponding to eq .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) gradually grow , and the scaling range finally becomes @xmath157 .
the fluctuation in @xmath158 is large , which may be the finite - size effect .
particularly , the time development of @xmath22 is large .
we note the calculation of the bogoliubov - wave distribution @xmath19 . as pointed out in sec .
iv b , in previous works @xcite , the condensate function is assumed to be @xmath134 .
when we use this condensate function and calculate the bogoliubov - wave distribution , the spectrum @xmath22 does not show a @xmath159 power law and the deviation from this law is much larger .
when we use the condensate function defined by eq .
( [ condensate ] ) , however , the spectrum @xmath22 shows good agreement with eq .
( [ b_sol2 ] ) .
therefore , it is important to take account of the dynamics of the condensate coupled with the fluctuation .
such a condensate has been called a quasicondensate @xcite , but the condensate dynamics of eq .
( [ gp_condensate ] ) has not been treated in previous studies . finally , we comment on the relation between the power exponents and the strength of the fluctuation . due to our analytical calculation , if the weak nonlinear condition is satisfied , the energy flux is small , and the power exponents of eqs .
( [ b_sol2 ] ) , ( [ b_w_rel3 ] ) , and ( [ w_d_p_rel4 ] ) can appear , although we do not numerically confirm the appearance of the same power laws in the case of weaker fluctuation because it takes a much longer time to do the numerical calculation . on the other hand , if the fluctuation is strong , the energy flux is large , and other power exponents may appear .
such a strong wt is the future work .
in this section , we derive the three power exponents for @xmath22 , @xmath20 , and @xmath21 . in sec . v a
, we detail the difference between the power exponents in previous results and ours . in sec . v b ,
we propose some conditions to observe kolmogorov turbulence in atomic becs and comment on the turbulence theory beyond the weak wt one .
we consider that our result for @xmath21 may be the key to achieve this observation . in sec .
v c , we discuss an inconsistency of our analytical calculation in sec .
iii . in a previous study @xcite , the spectrum of the macroscopic wave function
was suggested to obey a @xmath0 power law in a three - dimensional system .
however , our result has another @xmath1 power law . to discuss the reason for this difference , we review the derivation of the @xmath0 power law in this previous paper @xcite . in @xcite ,
zahkarov and nazarenko studied bogoliubov - wave turbulence using the two canonical variables @xmath160 and @xmath161 with @xmath162 . to study the fluctuation around the strong condensate
, they took expressions @xmath163 and @xmath164 , where the quantity @xmath165 corresponds to the condensate and @xmath166 and @xmath167 correspond to the fluctuations . by using the following canonical transformation for these canonical variables
, they introduced new canonical variables @xmath168 , diagonalizing a one - body hamiltonian as @xmath169 the canonical variables @xmath168 were defined by @xmath170 @xmath171 with the fourier components @xmath172 $ ] and @xmath173 $ ] . here , note that @xmath174 is different from @xmath50 of eq .
( [ hamiltonian4 ] ) .
then , they derived the @xmath0 power law in the bogoliubov - energy spectrum defined by @xmath175 the previous paper @xcite makes use of this power law to derive the power exponent of the spectrum for the macroscopic wave function . in this previous paper @xmath176
is estimated as @xmath177 with @xmath178 .
then , in the low - wave - number region , @xmath179 is approximately satisfied , so that they derived the @xmath0 power law in the spectrum @xmath20 for the macroscopic wave function .
however , owing to the weak fluctuation , the macroscopic wave function @xmath3 can be expanded as @xmath180 by applying the fourier transformation to eq .
( [ comp1 ] ) , we obtain @xmath181 as a result , in the low - wave - number region , the approximation used in weak wt theory gives @xmath182 which obviously leads to the @xmath1 power law in @xmath20 because @xmath183 .
thus , the derivation of the @xmath0 power law for @xmath20 in the previous work @xcite seems not to be correct . in atomic bec experiments
, observations can be made of the density distribution , which accesses the spectrum @xmath21 .
this can be the key to confirming the presence of kolmogorov turbulence ; if we can observe the @xmath0 power law in this spectrum , then kolmogorov turbulence is present .
we discuss three points related to observing the @xmath0 power law in @xmath21 .
the first point is that one must take a wide scaling region ; that is , one must prepare an atomic bec that is much larger than the coherence length .
the @xmath0 power law appears in the low - wave - number region @xmath184 , but , in the trapped system , there is a limit on the system size of the order of the thomas - fermi radius @xmath185 .
this scale can be the lower limit of the scaling region for the @xmath0 power law .
thus , the radius @xmath185 should be at least a factor of 10 larger than @xmath186 .
the second point concerns how to excite the system weakly in the low - wave - number region .
the @xmath0 power law is valid under the weak nonlinearity condition @xmath153 , so that fluctuations of both density and phase should be very weak . in our calculation , the density fluctuation is about @xmath187 as shown in fig .
however , this may be too weak to observe .
we must investigate the spectrum dependence on the strength of the fluctuation , for instance , the condition at which the prediction of weak wt theory is broken ; this , however , is a subject of future work . also , it is necessary to excite a large - scale modulation because the @xmath0 power law results from the direct energy cascade .
thus , candidates for the excitation are considered to be repulsive or attractive gaussian potentials moving slower than the sound velocity or a shaking trapping potential with fixed frequency . however
, the excitation mode induced by these methods should depend on size , strength , and the velocity of the gaussian potential or frequency and displacement of the shaking , etc .
understanding the excitation methods is essential for observing kolmogorov turbulence in atomic becs and should be numerically investigated hereafter .
the third point is that one must take surface effects into consideration . in a three - dimensional system ,
we can obtain the density distribution integrated along the incident direction of the probe light , so that this distribution contains information of the surface profile as well as the bulk .
since our result should be applicable to the bulk of the trapped bec , the surface effect may alter the power law of eq .
( [ w_d_p_rel4 ] ) .
if the system size is much larger than the damping length of the surface wave into the bulk @xcite , this power law is expected to be relevant . finally
, we comment on turbulence theory beyond weak wt theory .
recently , yoshida and arimitsu have studied strong turbulence with the spectral closure approximation , deriving the some power exponents @xcite .
these power exponents have not been confirmed numerically .
these power laws may be another key to confirmation of kolomogorov turbulence in atomic becs because , in this theory , the fluctuation is large and it is easy to observe . in sec .
iii and appendix a , we apply weak wt theory to eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) , obtaining eq .
( [ kin_equ1 ] ) . however , this derivation has an inconsistency for the perturbative expansion in eqs .
( [ gp_condensate ] ) and ( [ gp_fluctuation2 ] ) . in this expansion , we keep the second order of the fluctuation @xmath188 in both equations , which is not consistent because the right - hand sides of eqs .
( [ gp_condensate ] ) and ( [ gp_fluctuation2 ] ) are the first and second order of smallness , respectively . to do a reliable expansion
, we must take the third order of @xmath188 in eq .
( [ gp_fluctuation2 ] ) , but this calculation is very difficult .
our numerical calculation exhibits that the condensate dynamics is different from @xmath46 , so that we must take the nonlinear dynamics for @xmath29 . in this paper , although our calculation contains the inconsistency for the perturbative expansion , we treat the coupled nonlinear dynamics for the condensate and the fluctuation . thus , our theory for this coupled dynamics is insufficient .
the construction of the improved theory is the future work . in spite of this inconsistency , the analytical results exhibit good agreement with the numerical results . as another possibility
, our numerical calculation seems not to be substantially satisfied with the weak nonlinear condition .
however , if we use the weaker fluctuation , the computational time for the formation of the power law is much longer , which can not be confirmed at present .
we have theoretically and numerically studied bogoliubov - wave turbulence in a uniform bec at zero temperature by using the gp equation .
our application of weak wt theory to the gp equation leads to the three power exponents @xmath159 , @xmath1 , and @xmath189 corresponding to the spectra for the bogoliubov - wave distribution @xmath22 , the macroscopic wave function @xmath20 , and the density distribution @xmath21 . by numerically calculating the gp equation , we have confirmed the good agreement with these power exponents . in our analytical and numerical calculation , we found that taking account of the condensate dynamics coupled with the fluctuation , an aspect not treated in previous works @xcite , was important .
the authors thank s. nazarenko for pointing out the importance of the dimension in the acoustic and bogoliubov - wave turbulence .
k.f . was supported by a grant - in - aid for jsps fellows ( grant no .
mt was supported by jsps kakenhi grant no .
26400366 and mext kakenhi `` fluctuation @xmath190 structure '' grant no .
we apply weak wt theory in @xcite to eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) , deriving eq .
( [ kin_equ1 ] ) . in this theory ,
the separation of time scale is important .
there are two time scales , which are a linear time scale @xmath191 and a nonlinear one @xmath192 . in weak wt ,
the nonlinear term is very weak , which leads to the long nonlinear time scale @xmath193 . in weak wt theory
, we make use of this time - scale separation , filtering out the fast linear oscillation . for this purpose ,
let us introduce auxiliary intermediate time @xmath194 satisfying @xmath195 .
in the following , we solve eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) with a formal perturbative expansion , finding the canonical variables @xmath19 at @xmath196 . before the perturbative expansion , we rewrite eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) with interaction variables @xmath197 defined by @xmath198 @xmath199 where we introduce the small parameter @xmath200 in order to elucidate the power of smallness of each term in the perturbative expansion .
then , eqs .
( [ density_equ ] ) and ( [ bogo_equ ] ) become @xmath201 , \label{ap1}\end{aligned}\ ] ] @xmath202 with @xmath203 $ ] and @xmath204 $ ] .
we do not show the expression of @xmath205 because it is not important . in this calculation , as noted in sec .
iii c , the terms with @xmath60 are neglected .
we calculate the canonical variable @xmath197 and the condensate density @xmath69 at @xmath196 up to the second order of @xmath200 by using the perturbative expansion : @xmath206 @xmath207 from eqs .
( [ ap2 ] ) and ( [ ap4 ] ) , the expansion of @xmath69 is given by @xmath208 the function @xmath209 such as @xmath59 , @xmath210 , and @xmath211 has @xmath69 as a variable , being expanded in the following : @xmath212 thus , the second order of the condensate density @xmath213 can be neglected in eq .
( [ ap1 ] ) because such terms lead to the third order of @xmath200 in eq .
( [ ap1 ] ) .
therefore , the time dependence of @xmath69 vanishes in the application of weak wt theory , so that we can use the conventional derivation of the kinetic equation in @xcite , obtaining eq .
( [ kin_equ1 ] ) .
the derivation of eq . ( [ interaction4 ] ) is described . in the low - wave - number region ,
the momentum and energy resonant conditions of the bogoliubov waves give the particular class of the wave number , where the wave vectors are nearly parallel to each other because of the linear dispersion @xcite .
then , the resonant condition @xmath214 becomes @xmath215 @xmath216 where the angle @xmath217 formed by the two vectors @xmath218 and @xmath219 is small .
taking the second order of the small quantities , we obtain the angle @xmath220 @xcite , which gives @xmath221 here we use the condition @xmath222 because of the resonant condition .
thus , the effective interaction between the bogoliubov waves can be calculated by @xmath223 of eq .
( [ interaction3 ] ) . as a result
, we obtain @xmath224
we show the derivation of eq .
( [ bogkin1 ] ) from eq .
( [ kin_equ1 ] ) .
we assume the isotropy of @xmath83 in the weak wt , integrating the @xmath219 part and taking the leading terms , which leads to @xmath225 @xmath226,\end{aligned}\ ] ] with @xmath227 , @xmath228 , and @xmath229 .
finally , we can perform the integration by substitution with @xmath230 and @xmath231 ( @xmath232 ) , obtaining eq .
( [ bogkin1 ] ) .
the locality condition of the power law is derived . in this derivation
, we consider the convergence condition of collisional integral in eq .
( [ bogkin1 ] ) .
the power law is assumed to be @xmath113 with the positive power exponent @xmath5 .
then , from eq .
( [ bogkin1 ] ) , we find that the collisional integral may diverge at three points @xmath233 .
the point @xmath234 means the @xmath235 , so that the convergence condition at @xmath234 is the same as that at @xmath236 because of the exchange symmetry ( @xmath237 ) of eq .
( [ kin_equ1 ] ) .
thus , in the following , the convergence conditions at @xmath238 are described . in the large @xmath239 region ,
the dangerous terms in the collisional integral are the second and third ones of eq .
( [ bogkin1 ] ) , which is approximately expressed by @xmath240 \nonumber \\ & \simeq & 4\int ^{\infty } dk_1 kk_1^{4}n(k ) \frac{\partial n(k_1)}{\partial k_1 } \nonumber \\ & \propto & \int ^{\infty } dk_1 k_1^{3-m}.\end{aligned}\ ] ] thus , the convergence condition at @xmath241 is @xmath242 . on the other hand , in the small @xmath239 region ,
the dangerous terms appear in the first and third ones of eq .
( [ bogkin1 ] ) .
these terms become @xmath243 \nonumber \\ & \simeq & \int _ 0 dk_1 k_1^{4}n(k_1 ) \frac{\partial^2 n(k)}{\partial k^2 } \nonumber \\ & \propto & \int_0 dk_1 k_1^{4-m } , \end{aligned}\ ] ] which shows that the convergence condition at @xmath236 is @xmath244 .
99 p. a. davidson , _ turbulence : an introduction for scientists and engineers _ ( oxford university press , oxford , 2004 ) .
u. frisch , _ turbulence _
( cambridge university press , cambridge , 1995 ) . , edited by w. p. halperin and m. tsubota ( elsevier , amsterdam , 2008 ) , vol .
w. f. vinen , j. low temp . phys . * 161 * , 419 ( 2010 ) . l. skrbek and k. r. sreenivasan , phys .
fluids * 24 * , 011301 ( 2012 ) .
a. c. white , b. p. anderson , and v. s. bagnato , proc .
* 111 * , 4719 ( 2013 ) c. f. barenghi , l. skrbek , and k. r. sreenivasan , proc .
* 111 * , 4647 ( 2014 ) .
w. f. vinen , proc .
a * 240 * , 114 ( 1957 ) ; * 240 * , 128 ( 1957 ) ; * 242 * , 493 ( 1957 ) ; * 243 * , 400 ( 1958 )
. j. t. tough , in _ progress in low temperature physics _ , edited by d. f. brewer ( north - holland , amsterdam , 1982 ) , vol .
c. nore , m. abid , and m. e. brachet , phys .
78 * , 3896 ( 1997 ) ; phys . fluids * 9 * , 2644 ( 1997 ) .
m. kobayashi and m. tsubota , phys .
lett . * 94 * , 065302 ( 2005 ) ; j. phys .
* 74 * , 3248 ( 2005 ) .
j. mauer and p. tabeling , europhys .
lett . * 43 * , 29 ( 1998 ) .
e. kozik and b. svistunov , phys .
* 92 * , 035301 ( 2004 ) .
v. s. lvov and s. nazarenko , j. low . temp .
. lett . * 91 * , 428 ( 2008 ) .
v. s. lvov , s. v. nazarenko , and o. rudenko , phys .
b * 76 * , 024520 ( 2007 ) .
m. s. paoletti , r. b. fiorito , k. r. sreenivasan , and d. p. lathrop , j. phys .
. jpn . * 77 * , 111007 ( 2008 ) m. s. paoletti , m. e. fisher , k. r. sreenivasan , and d. p. lathrop , phys . rev
* 101 * , 154501 ( 2008 ) .
t. w. neely , e. c. samson , a. s. bradley , m. j. davis , and b. p. anderson , phys .
lett . * 104 * , 160401 ( 2010 ) . c. j. pethick and h. smith , _ bose - einstein condensation in dilute gases _ , 2nd ed .
( cambridge university press , cambridge , 2008 ) . k. kasamatsu , m. tsubota , and m. ueda , int
b * 19 * , 1835 ( 2005 ) .
y. kawaguchi and m. ueda , phys . rep . * 520 * , 253 ( 2013 ) .
d. m. stamper - kurn and m. ueda , rev .
mod . phys . * 85 * , 1191 ( 2013 ) .
s. gautam and d. angom , phys .
a * 81 * , 053616 ( 2010 ) .
k. sasaki , n. suzuki , d. akamatsu , and h. saito , phys .
a * 80 * , 063611 .
h. takeuchi , n. suzuki , k. kasamatsu , h. saito , and m. tsubota , phys .
b * 81 * , 094517 ( 2010 ) .
m. karl , b. nowak , and t. gasenzer , phys .
a * 88 * , 063615 ( 2013 ) .
d. kobyakov , a. bezett , e. lundh , m. marklund , and v. bychkov , phys .
rev . a * 89 * , 013631 ( 2014 ) .
b. villase@xmath246or , r. zamora - zamora , d. bernal , and v. romero - roch@xmath247n , phys .
rev . a * 89 * , 033611 ( 2014 ) .
k. fujimoto and m. tsubota , phys .
a * 90 * , 013629 ( 2014 ) .
n. g. parker and c. s. adams , phys .
95 * , 145301 ( 2005 ) .
m. kobayashi and m. tsubota , phys .
a * 76 * , 045603 ( 2007 ) .
horng , c .- h .
hsueh , s .- w .
su , y .- m .
kao , and s .- c .
gou , phys .
a * 80 * , 023618 ( 2009 ) .
a. s. bradley and b. p. anderson , phys .
x * 2 * , 041001 ( 2012 ) .
m. t. reeves , t. p. billam , b. p. anderson , and a. s. bradley , phys .
* 110 * 104501 ( 2013 ) .
a. c. white , c. f. barenghi , and n. p. proukakis , phys .
a * 86 * , 013635 ( 2012 ) .
t. simula , m. j. davis , and k. helmerson , phys .
113 * , 165302 ( 2014 ) .
g. w. stagg , a. j. allen , n. g. parker , and c. f. barenghi , phys .
rev . a * 91 * , 013612 ( 2015 ) .
e. a. l. henn , j. a. seman , g. roati , k. m. f. magalhaes , and v. s. bagnato , phys .
lett . * 103 * , 045301 ( 2009 ) .
m. t. reeves , b. p. anderson , and a. s. bradley , phys .
a * 86 * , 053621 ( 2012 ) .
t. w. neely , a. s. bradley , e. c. samson , s. j. rooney , e. m. wright , k. j. h. law , r. carretero - gonz@xmath248lez , p. g. kevrekidis , m. j. davis , and b. p. anderson , phys .
lett . * 111 * , 235301 ( 2013 ) .
w. j. kwon , g. moon , j. y. choi , s. w. seo , and y. i. shin , phys .
a * 90 * , 063627 ( 2014 ) v. e. zakharov , v. s. lvov , and g. falkovich , _ kolmogorov spectra of turbulence i : wave turbulence _
( springer , berlin , 1992 ) .
s. nazarenko , _ wave turbulence _ , lecture notes in physics vol .
825 ( springer , heidelberg , 2011 ) .
@xmath249 . falcon , c. laroche , and s. fauve , phys .
98 * , 094503 ( 2007 ) . v.
s. lutovinov and v. r. chechetkin , zh .
. fiz . * 76 * , 223 ( 1979 ) .
g. dring , c. josserand , and s. rica , phys .
lett . * 97 * , 025503 ( 2006 ) .
s. dyachenko , a. c. newell , a. pushkarev , and v. e. zakharov , physica d * 57 * , 96 ( 1992 ) .
v. zakharov and s. nazarenko , physica d * 201 * , 203 ( 2005 ) .
d. proment , s. nazarenko , and m. onorato , phys .
a * 80 * , 051603(r ) ( 2009 ) .
g. e. falkovich , zh .
teor . fiz .
* 93 * , 117 ( 1987 ) . v. s. lvov
lvov , a. c. newell , and v. zakharov , phys .
e , * 56 * , 390 ( 1997 ) .
m. kobayashi and m. tsubota , phys .
97 * 145301 ( 2006 ) .
s. nazarenko , m. onorato , and d. proment , phys .
a * 90 * , 013624 ( 2014 ) .
k. yoshida and t. arimitsu , j. phys .
a : math . theor . * 46 * , 335501 ( 2013 ) . | we theoretically and numerically study bogoliubov - wave turbulence in three - dimensional atomic bose - einstein condensates with the gross - pitaevskii equation , investigating three spectra for the macroscopic wave function , the density distribution , and the bogoliubov - wave distribution . in this turbulence
, bogoliubov waves play an important role in the behavior of these spectra , so that we call it bogoliubov - wave turbulence . in a previous study [ d. proment _ et al .
_ , phys .
rev .
a * 80 * , 051603(r ) ( 2009 ) ] , a @xmath0 power law in the spectrum for the macroscopic wave function was suggested by using weak wave turbulence theory , but we find that another @xmath1 power law appears in both theoretical and numerical calculations .
furthermore , we focus on the spectrum for the density distribution , which can be observed in experiments , discussing the possibility of experimental observation . through these analytical and numerical calculations ,
we also demonstrate that the previously neglected condensate dynamics induced by the bogoliubov waves is remarkably important . |
experiments searching for neutrinoless double beta ( @xmath0 ) decay require an extremely low background level in the region of interest around a few mev .
compton scattered @xmath1-particles , originating from radioactive decays in the proximity of the detectors , are an important background contribution at such energies . in high - purity germanium ( hpge ) experiments ,
these interactions are often identified and removed from the signal data set through pulse - shape analysis ( psa ) . in order to extract a half - life limit ,
the signal recognition efficiency has to be known . usually , experimentally obtained pulse - shape libraries with signal - like events are used to obtain the signal recognition efficiency . however , these evaluation libraries can have energy - deposition topologies and event - location distributions different to those of the signal searched for .
efficiencies obtained like this can be systematically different from the recognition efficiency for the real signal . furthermore , the evaluation libraries used to derive the recognition efficiencies often contain events of the wrong type , making a direct determination of the efficiencies impossible .
this paper presents investigations of the reproducibility and systematic uncertainties of the efficiencies of pulse - shape discrimination ( psd ) using artificial neural networks ( anns ) with libraries of simulated pulses .
the general idea of psd using anns is introduced and the sources of possible systematic effects are discussed . the simulations and the libraries used for the analysis are described as well as the anns and the procedures used to train them .
the stability of the method against initial conditions and ann topologies is investigated .
the focus is on the differences obtained in recognition efficiencies using different evaluation libraries and the associated systematic uncertainties .
the detection principle of semiconductor detectors is based on the creation and detection of electron hole pairs , i.e. charge carriers , when radiation interacts with the detector material .
charge - sensitive preamplifiers are commonly used to detect the drifting charge carriers in large volume hpge detectors .
the time structure of an event , the pulse - shape , is defined by the mirror charge signal induced on the electrodes as a function of time .
the pulse length is given by the time needed to fully collect the charges on the electrodes .
see e.g. @xcite for a detailed description of the pulse creation process . for photons in the mev range ,
the dominant interaction process is compton scattering .
a photon with an energy of one mev has a mean free path of @xmath23 cm in germanium .
thus , photon - induced events with energies of about 2mev are mostly composed of several energy deposits within a hpge detector , separated by a few centimeters .
these background - like events are referred to as multi - site events ( mse ) .
in contrast , electrons with the same energy have a range of the order of millimeters and deposit their kinetic energy `` locally '' .
signal - like events of this kind are referred to as single - site events ( sse ) .
note that in reality there also exists `` signal - like '' background , i.e. background events that have an indistinguishable event topology , such as the irreducible background from @xmath3 decay .
the two electrons emitted in 0@xmath4 decay result predominantly in sses . due to bremsstrahlung ,
a fraction of a few % of the 0@xmath4-decay events become mses @xcite .
events identified as mse in the energy region of interest are rejected as background .
methods to distinguish between sses and mses in hpge detectors using anns were developed previously @xcite . in most previous works , events from double escape peaks ( dep ) and full energy peaks ( fep ) were used to create training libraries of signal - like and background - like events , respectively .
these were obtained from calibration data for which sources such as @xmath5th or @xmath6co were used . the ann efficiencies to correctly identify events
are also typically extracted using evaluation libraries from calibration measurements .
the efficiencies of psd methods are not necessarily homogeneous throughout the detector volume . for a realistic evaluation
, the spatial distribution of the events in a given test library has to be taken into account .
especially , dep events will exhibit a non - uniformity in event location distribution due to the topology of the events .
if pair production occurs in a coaxial hpge at high radii , @xmath7 , and height , @xmath8 , i.e. close to the extreme boundaries , the probability for the two 511kev @xmath1-particles to escape is the highest . hence , libraries of dep events have a higher event location density in these parts of the detector ( see section [ sec : libraries ] and fig .
[ fig : psa : event_distr ] ) . on the other hand ,
signal events due to @xmath0 decay ( but also `` signal - like '' background events due to @xmath3 decay ) are expected to be homogeneously distributed . using a library with an event location distribution different from the one expected for
the signal can lead to systematic biases .
the main scope of this work is to address this issue and estimate the uncertainties on the sse recognition efficiency evaluation arising from the use of different training and evaluation sets .
in order to quantify the uncertainties on the ann event topology recognition efficiencies , simulations are used .
the signal ( background ) recognition efficiency @xmath9 ( @xmath10 ) of any psd method is defined as the probability that the method correctly identifies an sse ( mse ) from an event - library containing only sses ( mses ) .
realistic sse and mse pulse - shape libraries always contain events of both classes .
hence , the ann method applied to a library of predominantly sse or mse pulses will result in a survival probability @xmath11 or rejection probability @xmath12 , defined as the fraction of pulses in the library that are classified as sse or mse , respectively : @xmath13 where @xmath14 and @xmath15 are the fraction of sses and mses in the sse - library , respectively , and @xmath16 and @xmath17 are the fraction of mses and sses in the mse - library , respectively . using simulated pulses idealized libraries with @xmath18 and @xmath19 can be created .
these libraries can be used to determine @xmath9 and @xmath10 directly , as for this case @xmath20 and @xmath21 ( see equ .
[ equ:1 ] ) . in order to quantify the effect of non - homogeneous event - location distributions , @xmath9 for anns obtained with training libraries with inhomogeneous event - location distributions
are compared to those obtained from libraries with a homogeneous event - location distribution .
the stability of the method is verified by training and evaluating a set of anns with different initial weights of the ann synapses and with training libraries of different sizes .
finally , the influence of the number of hidden layers and the number of neurons in the ann on @xmath9 , is investigated .
signal recognition efficiencies obtained using evaluation libraries with event location distributions as expected and different from the signal are then compared .
true - coaxial hpge detectors are considered in this paper .
they have a simple radial electric field and , thus , have relatively simple pulse shapes .
consequently , pulse - shapes of this type of detectors have lower systematic uncertainty due to smaller uncertainties in the field calculations compared to detectors with more complex geometries .
this makes this type of detectors interesting for this analysis .
hpge detectors for low background experiments typically have a radius , @xmath22 , and a height of a few cm axis pointing upwards . in cartesian
coordinates , the @xmath23- and @xmath24- axes coincide with the crystallographic @xmath25 axes , while the @xmath8 axis coincides with the crystallographic @xmath26 axis . ] .
the simulated n - type true - coaxial germanium detector has a height of 70 mm and @xmath22=37.5 mm with the diameter of the borehole being 10 mm .
the dead layer due to the n+ contact ( outer surface ) is less than 1@xmath27 m , while the dead layer due to the p+ contact is 0.5 mm .
the simulated geometry describes an existing true - coaxial 18-fold segmented n - type hpge developed as a prototype detector @xcite for the gerda experiment @xcite .
photon and electron interactions for different libraries were simulated within the mage framework @xcite , based on geant4 @xcite .
pulse shapes were simulated for the core electrode . whenever individual energy deposits within one event were separated by less than 0.1 mm they were combined .
pulse shapes for the combined energy deposits were simulated using pre - calculated electric and weighting fields using the pulse - shape simulation package described in @xcite .
the charge collection efficiency is either zero or one .
charge cloud diffusion and self repulsion effects are not taken into account in the simulation .
the drift path anisotropy originating from the axis effect , i.e. the dependence of mobilities on the axis orientation , is accounted for in all simulations .
the number of grid points for the electric- and weighting - field calculations was @xmath28 .
the electrically active impurities were assumed to be homogeneous within the detector , with a density of @xmath29 .
the length of the simulated pulses is 1@xmath27s .
the step frequency of the simulation is 1125mhz , a multiple of 75mhz to which the pulses were resampled to take the effects of a typical daq into account . above 1ghz ,
the step frequency is sufficient to correctly describe trajectories @xcite .
the amplifier rc - integration constant was set to 20ns , corresponding to a bandwidth of about 10mhz , while the amplifier decay time was set to 50@xmath27s .
each individual pulse shape was convoluted with gaussian noise , @xmath30kev .
the results presented in this work do not change when simulated pulses with no noise are used , i.e. the efficiencies obtained are within the uncertainties quoted in the following .
sse and mse pulses take on a wide variety of shapes as shown in fig .
[ fig : sample_pulse ] .
it is not trivial to interpret the pulse shapes and distinguish between sses and mses without an involved quantitative analysis .
the pulse length , @xmath31 , is between 160 and 500ns @xcite , where @xmath31 is defined as the time in which the pulse increases from 10% to 90% of its amplitude .
this part of the pulse contains the relevant information regarding the event topology .
training and evaluation libraries with independent pulses were created .
the simulated libraries are listed in table [ tab : libraries ] .
the dep , 2@xmath4 and 0@xmath4 event - libraries were created with and without a realistic admixture of mses due to bremsstrahlung and compton - scattered @xmath1-particles .
all mse libraries were simulated for the 1620kev fep , corresponding to a @xmath5th source , typically used for calibration .
the notation of _ no comp _ & _ brems _ is used for sse libraries in which all events with compton scattering or hard bremsstrahlung were removed .
mse libraries that contain only events which have at least one energy deposition due to compton scattering or hard bremsstrahlung in the detector and thus have at least two distinct energy deposits are marked as _ comp _ & _ brems only_. in order to obtain a clean mse library it was required that r@xmath32 , the radius within which 90% of the deposited energy was contained @xcite , is larger than 2 mm .
this ensures that all events have at least two energy deposits that are at least 2 mm apart . to indicate the origin of incoming photons , the last column in table [ tab : libraries ] lists either `` top '' , `` side '' or `` homog '' .
this means that the photons were simulated to come from either the @xmath33- or @xmath34-plane for `` top '' and `` side '' , respectively .
their origins are homogeneously distributed on these planes with their momentum perpendicular to the plane of origin .
the planes are located 17.5 cm from the center of the detector and their area is sufficiently large to cover the detector .
sse libraries with homogeneous event location distributions within the detector volume are listed as `` homog '' .
for _ dep clean _ , 2.6mev photons were forced to make pair creation with the event vertices homogeneously distributed within the detector .
each training and evaluation library contains between 7.000 and 20.000 simulated pulses .
lcccc library & energy & processes & source location + + _ dep top _ & ( 1593@xmath355)kev & _ no comp _ & _ brems _ & top + _ dep side _ & ( 1593@xmath355)kev & _ no comp _ & _ brems _ & side + _ dep real _ & ( 1593@xmath355)kev & all processes & side + _ dep clean _ & ( 1593@xmath355)kev & _ no comp _ & _ brems _ & homog . + _ 0@xmath4
real _ & ( 2039@xmath355)kev & all processes & homog .
+ _ 0@xmath4 clean _ & ( 2039@xmath355)kev & _ no comp _ & _ brems _ & homog . + _
2@xmath4 real _ & 450kev540kev & all processes & homog . + _
2@xmath4 clean _ & 1000kev1450kev & _ no comp _ & _ brems _ & homog . + + _ fep top _ &
( 1620@xmath355)kev & _ comp _ & _ brems only _ & top + _
fep all _ & ( 1620@xmath355)kev & all processes & top + _ fep side _ & ( 1620@xmath355)kev & _ comp _ & _ brems only _ & side + _ fep clean _ & ( 1620@xmath355)kev & r@xmath362 mm , _ comp _ & _ brems only _ & top + the radial distributions of the energy barycenters , defined as the energy - weighted mean radial position of the energy deposit , of individual events for sse libraries containing no mses are shown in fig .
[ fig : psa : event_distr ] .
top , middle and bottom refer to events contained in the upper , middle and lower third of the detector , respectively .
these three volumes are equal .
the barycenter of an individual event corresponds approximately to the position of the interaction / decay .
for the _ dep clean _
library , where _ clean _ is used here and below to identify libraries with no compton or bremsstrahlung interactions , it is flat as a function of @xmath7 and equivalent to the distribution of the _ 2@xmath4 real _ library . _ real _ is used to indicate libraries which contain all processes , i.e. including compton scattering and bremsstrahlung .
dep side _ and _ dep top _ libraries have inhomogeneous event - location distributions , events being located with a higher probability at high @xmath7 close to the bottom and top of the detector since for these parts of the detector it is more likely for the two back to back 511kev photons to escape the detector .
_ side _ and _ top _ indicate the location of the source with respect to the detector .
the libraries listed in table [ tab : libraries ] were used to create five different ann training and evaluation sets each . they are listed in table [ tab : training_sets ] , showing the combinations of sse and mse libraries .
lcc & sse library & mse library + & _ dep side _ & _ fep side _ + _ set ii - real 2@xmath4 _ & _ 2@xmath4 real _ & _ fep top _ + _ set iii - hom dep _ & _ dep clean _ & _ fep top _ + _ set iv - top dep _
& _ dep top _ & _ fep top _ + _ set v - clean 0@xmath4 _ & _ 0@xmath4 clean _ & _ fep top _
+ the anns used in this analysis were built using the tmultilayerperceptron ( tmlp ) within the root framework @xcite . only the part of the pulse containing the relevant information on the event topology is used by the anns .
the pulses in the considered detector are maximally around 500ns long . in total , 40 time steps , corresponding to 530ns , were used .
the center of the resulting trace was chosen to be the point where the pulse reaches 50% of its amplitude .
the amplitude of each pulse was normalized to unity .
the anns are composed of 40 input neurons , one hidden layer with the same number of neurons and an output layer with only one neuron .
the anns were trained using the broyden - fletcher - goldfarb - shanno learning method @xcite .
background - like mses were assigned an ann output , @xmath37 , of 0 and signal - like sses were assigned an @xmath37 of 1 .
libraries of the same size for mses and sses were used . for a trained network ,
@xmath37 should be close to 1 for sses and close to 0 for mses . for each individual ann events
are classified as sse if @xmath38 , where @xmath39 is a parameter that has to be optimized . the rejection probability @xmath40 represents the fraction of events from an mse dominated library , fep in this case , rejected by the cut @xmath41 .
the survival probability @xmath42 represents the fraction of events from a sse dominated library ( dep , @xmath3 or @xmath0 ) kept with @xmath38 .
the cut value @xmath43 is chosen for each individual ann to maximize the quantity @xmath44 of the corresponding evaluation set used .
this ensures that the highest e and r are obtained at the same time . the solid histogram in fig .
[ fig : sample_spectrum ] shows the simulated energy spectrum for events contained in the _ dep real _ and _ fep all _ libraries ( table [ tab : libraries ] ) . the fep is significantly reduced while the dep remains almost untouched .
the survival probability @xmath42 for 0@xmath4 and dep events is given by the ratio of the peak areas after and before the ann rejection .
the areas are determined by fitting a gaussian plus constant background to the spectra . in figures [ fig : psa
: training_a ] and [ fig : psa : training_b ] , the @xmath37 distribution for an ann trained with mses and sses from training _ sets i _ and _ ii _ are shown .
a clear separation between the @xmath37 distributions of the mse and sse libraries is visible .
figures [ fig : psa : training_c ] and [ fig : psa : training_d ] show @xmath42 , @xmath40 and @xmath45 ) for training _ sets i _ and _ ii _ , respectively .
the vertical line represents @xmath46 .
the @xmath47 , @xmath48 , @xmath49 and @xmath50 values are called @xmath11 , @xmath12 , @xmath51 and @xmath9 , respectively , in the following .
these variables are summarized in table [ tab : variables ] . for libraries with purely sse or mse events , for which @xmath18 and @xmath52 , @xmath11 and @xmath12 coincide with @xmath9 and @xmath10 , respectively ( see eq .
[ equ:1 ] ) .
note that the anns with the optimized @xmath46 as described here are later used for efficiency and uncertainty evaluations .
clcc survival probability & & @xmath53 + rejection probability & & @xmath54 + signal recognition efficiency & & @xmath55 + background recognition efficiency & & @xmath56 + background reduction power & & @xmath57 + statistical uncertainties quoted in the following are derived from the statistical fluctuations expected due to the limited number of simulated events and events surviving the selection .
the reproducibility of @xmath9 was investigated by training five anns with the same ann topology .
the same training samples were used but the initial weights of the individual synapses of the untrained ann were different in each case . also the order in which individual pulses from the training sets were chosen for the iterative training was different for each ann .
the fluctuations of @xmath9 between the different anns are of the order of 1% of the value of @xmath9 , the rms of the distribution of @xmath9 is taken as its systematic uncertainty .
this systematic uncertainty only describes within which precision efficiencies are reproducible .
they are not to be confused with systematic uncertainties related to pulse shape simulation .
two groups of five anns , each with a different number of neurons in the hidden layer , were trained using the same training set ( _ set ii _ ) .
the value of @xmath9 for the default ann with 40 hidden neurons was @xmath58 ( stat . )
@xmath59 ( syst . ) , while an ann with 40 input neurons and one hidden layer with 10 neurons had a recognition efficiency @xmath60 ( stat . ) @xmath61 ( syst . ) .
this is not significantly worse .
five anns with three hidden layers with 40 neurons each were trained and have @xmath62 ( stat . )
@xmath63 ( syst . ) .
this is not a significant improvement with respect to the default ann .
the corresponding @xmath51 values are ( @xmath64 ) , ( @xmath65 ) and ( @xmath66 ) for the default network , the network with one hidden layer of 10 neurons and the network with three hidden layers , respectively . in summary , the variation of @xmath9 due to the choice of the topology of the network is @xmath67 . in the following ,
the default network with one hidden layer with 40 neurons is used and the variation due to the topology is not considered in the following uncertainties .
anns were trained with the training sets listed in table [ tab : training_sets ] .
the trained anns were applied to the libraries @xmath68 and _ fep clean _ , containing purely sses and mses . in this case ,
@xmath18 and @xmath52 , respectively .
hence , @xmath69 = @xmath9 and @xmath70 = @xmath10 for a clean library ( see eq . [ equ:1 ] ) .
the resulting @xmath9 , @xmath10 and @xmath51 values are given in table [ tab : efficiencies ] .
lccc training set & @xmath9 ( _ @xmath0 clean _ ) & @xmath10 ( _ fep clean _ ) & @xmath51 + _ set i inhom dep _ & [email protected] & [email protected] & 0.904 + _ set ii real 2@xmath4 _ & [email protected] & [email protected] & 0.917 + _ set iii
hom dep _ & [email protected] & [email protected] & 0.924 + _ set iv
top dep _ & [email protected] & [email protected] & 0.904 + _ set v
clean 0@xmath4_&[email protected] & [email protected] & 0.922 + the highest values for @xmath51 were obtained with anns trained with sse samples with homogeneous event - location distributions .
the @xmath51 values for anns trained with inhomogeneous samples are by approximately 0.02 lower .
the variation on @xmath9 is up to 0.06 and hence more pronounced than on @xmath10 ( @xmath710.03 ) .
the variations due to sse libraries with different event - location distributions used for the ann training are significantly bigger than the fluctuations due to changes of the ann initial conditions .
the survival probabilities @xmath11 obtained with the trained and @xmath46 optimized anns were evaluated on the sse libraries _
@xmath3 real _ , _
@xmath0 real _ and _ dep real _ according to the method explained in sec . [ sec : training ] .
the results are listed in table [ tab : survival_prob ] for training @xmath73 , @xmath74 and @xmath75 .
cccc ( -4,1)clean @xmath76 set bla bla & & & + _ 0@xmath4 real _ & [email protected] & [email protected] & [email protected] + & [email protected] & [email protected] & [email protected] + _ dep real _ & [email protected] & [email protected] & [email protected] + @xmath84 & ( [email protected])% & ( [email protected])% & ( [email protected])% + @xmath88 & ( [email protected])% & ( [email protected])% & ( [email protected])% + for 0@xmath4 events in the energy interval @xmath92 kev , @xmath93 values of ( [email protected] ) and ( [email protected] ) were obtained with the ann trained with _ set ii _ and _ set i _ , respectively , where the statistical and systematic uncertainties were added in quadrature . for the ann trained on _ set i _ , @xmath11 is lower than for _ sets ii _ and _ v _ , as expected from the lower @xmath9 for this training set ( see table [ tab : efficiencies ] ) .
the realistic signal - like libraries also contain a significant amount of mses .
this explains why the obtained @xmath11 values for 0@xmath4 are significantly different from the @xmath9 listed in table [ tab : efficiencies ] .
as the amount of wrong type of events in event libraries depends on geometry and energy this also implies that @xmath11 by itself is not a precise quantity to compare psd methods even if @xmath12 is also considered . the position distribution of the rejected events inside the detector , i.e. the position dependence of the signal recognition efficiency was studied .
in fig . [
fig : psa : xyaver ] , the location dependence of the mean value of the @xmath37 output inside the detector is depicted for the sses from the _
0@xmath4 clean _ library .
regions where the average @xmath37 output is lower than @xmath94 are seen as blue areas . in these regions ,
sses are systematically rejected .
the fraction of the volume where the sses of the _
0@xmath4 clean _ library are more likely to be rejected than to be accepted as sse is ( [email protected])% for an ann trained with the sse sample with inhomogeneous event - location distribution _ set i_. for the anns trained with _ sets ii _ and _ v _ , the affected volume is reduced to ( [email protected])% and ( [email protected])% , respectively @xcite .
using an ann training set with similar event - location distribution as for the evaluation set decreases the effect of the systematic volume cut , however , it does not completely remove it . the symmetry in the patterns observed in fig .
[ fig : psa : xyaver ] seems to be connected to the crystallographic symmetry of the detector .
the axis dependence of the effect might be due to the dependence of the electron to hole mobility ratio on the position of the charge carriers with respect to the crystal axes ( see fig . 2 in @xcite ) .
affected zones appear close to the inner detector surface and in the middle of the bulk around @xmath95 mm .
the mechanism of pattern formation is , however , not understood .
the different event - location distributions for dep samples from calibration and 0@xmath4 signal events ( see fig .
[ fig : psa : event_distr ] ) was identified as the major source of systematic uncertainty for the approach of anns trained with dep sets . for 2@xmath4 training samples
, the different energy distribution leads to a different signal - to - noise ratio .
the @xmath9 values obtained for different sse evaluation libraries with anns trained with different training sets are listed in table [ tab : efficiency_e ] .
the signal recognition efficiencies @xmath9 of the different anns are within uncertainties the same for the different evaluation libraries with homogenous event - location distribution .
this demonstrates that the normalization of the input to the ann makes the influence of the lower energy of events , down to 1mev , insignificant .
however , when @xmath9 is derived using the _ dep side _ set with realistic event - location distribution it is systematically overestimated .
there is a @xmath96(dep\ side - 0\nu\beta\beta)=(4.2^{+0.6}_{-0.9}\pm0.8)\%$ ] effect for the ann trained with an independent _ dep side _
sse training set . for anns trained with homogeneous samples
the effect is reduced but nevertheless significant with @xmath97(dep\ side - 0\nu\beta\beta)=(1.1^{+0.3}_{-0.4}\pm0.7)\%$ ] and @xmath98(dep\ side - 0\nu\beta\beta)=(1.4^{+0.4}_{-0.7}\pm0.9)\%$ ] .
note that the efficiencies obtained when training the network with the @xmath3 set are higher than the ones obtained using the @xmath0 set , the reason for which is unclear .
comparing the resulting @xmath9 with @xmath11 quoted in table [ tab : survival_prob ] shows that the additional admixture of mses to the evaluation libraries slightly reduces @xmath11 with respect to @xmath9 .
cccc ( -4,1)clean @xmath76 set bla bla & & & + _ @xmath0 clean _ & [email protected] & [email protected] & [email protected] + _ @xmath3 clean _ &
[email protected] & [email protected] & [email protected] + _ dep clean _ & [email protected] & [email protected] & [email protected] + _ dep side _ & [email protected] & [email protected] & [email protected] + @xmath103 ) & ( [email protected])% & ( [email protected])% & ( [email protected])% + @xmath106 ) & ( [email protected])% & ( [email protected]@xmath350.03)% & ( [email protected])% + @xmath108 ) & ( [email protected])% & ( [email protected])% & ( [email protected])% +
systematic effects on the determination of the signal recognition efficiency of pulse - shape - analysis using anns were investigated using pulse shape simulation .
the most important effect was found to be due to the event - location distribution of the evaluation libraries .
in contrast , the energy distribution of events in the training library was found to be irrelevant within reasonable limits .
the use of evaluation libraries with homogeneous event location distribution lead to reduced systematic uncertainties on the signal recognition efficiencies of the order of 1% . on the contrary signal recognition efficiencies of anns determined from dep libraries with inhomogeneous event location distributions
were found to be up to 5% too high , consistent with the systematic uncertainties derived in @xcite .
differences in the energy distribution of the events of the evaluation samples do not have a significant effect .
the different event - location distributions resulting from different positions of the calibration sources may result in variations of the ann signal recognition efficiency by up to 6% and the background discrimination power by 2% .
the signal detection efficiency of an ann depends on the location of the events inside a true - coaxial detector .
the efficiency is above 90% in most parts of the detector .
however , sses in the inner regions and in the center of the bulk are systematically misidentified . about 2% to 8% of the volume
is affected , depending on the homogeneity of the event - location distribution of the training set used .
using training sets with homogeneous sse location distribution reduces the affected regions but does not eliminate them completely .
the true - coaxial detectors assumed for these studies have particularly simple field configurations .
the effects on detectors with more complex field configurations will have to be studied very carefully .
pulse - shape discrimination with artificial neural networks is a useful tool to identify multi - site events .
it potentially increases the sensitivity of 0@xmath4 experiments like gerda @xcite . the usage of @xmath3 events for training and efficiency evaluation of the artificial neural networks is recommended . | a pulse - shape discrimination method based on artificial neural networks was applied to pulses simulated for different background , signal and signal - like interactions inside a germanium detector .
the simulated pulses were used to investigate variations of efficiencies as a function of used training set .
it is verified that neural networks are well - suited to identify background pulses in true - coaxial high - purity germanium detectors .
the systematic uncertainty on the signal recognition efficiency derived using signal - like evaluation samples from calibration measurements is estimated to be 5% .
this uncertainty is due to differences between signal and calibration samples . |
the light - front quark model ( lfqm ) @xcite , is based on the algebra of the generators of the lorentz - group in light - front dynamics @xcite .
the specific dynamical input is made through a parametrized form for the quark wave function . with specific choices of the parameters ,
the model describes numerous hadron properties , including form factors for @xmath3 1 gev@xmath4 ( see for example @xcite ) .
thus , this model establishes a phenomenological link between hadron properties and the wave function of the quark constituents that has been successful in many instances .
it is desirable to investigate a deeper theoretical connection between the lfqm wave function and qcd , not only to establish more firmly the dynamical basis of the lfqm but also with an eye to improve the general understanding and interconnections among various approaches . in our work
, we derive a connection between the lfqm and qcd light - cone wave functions by equating the matrix element @xmath5 in the two appproaches .
this leads to an expression for the light - cone qcd wave function in terms of the lfqm wave function .
some relevant questions were considered in @xcite .
particulary , in @xcite it was noted that the asymptotic behavior @xmath6 of the qcd light - cone wave function can not be reproduced within a constituent quark model ( cqm ) with equal - time wave functions @xcite @xmath7 , which can be represented in the form of the light - cone ( lc ) wave function by indentification ( see @xcite ) : q^2-m^2,_cqm(q^2 ) _ lc ( -m^2 ) , [ b1 ] where @xmath8 is the constituent mass .
the same conclusion is reached in the case of the lfqm @xcite .
therefore , we can not expect to obtain a good description of the qcd light - cone wave function in terms of the quark model near the end points @xmath1 . in the next section
we discuss the reason for this disagreement .
however , in most cases hadronic matrix elements calculated in the quark model are saturated at moderate @xmath0 , and the contribution of the region near the end points is small . in this case , one can expect that the lfqm can describe the light - cone wave functions for moderate @xmath0 . as mentioned above
, the lfqm is able to describe hadron matrix elements even for large values of momentum transfer : @xmath9 .
this means that the lfqm gives a good description of hadron matrix elements at small distances @xmath10 .
therefore , it is reasonable to use such a model for the evaluation of matrix elements of operators defined at a scale @xmath11 .
below , we will assume that the matrix elements ( which depend on the normalization point ) are defined at the scale @xmath12 .
in this paper we consider the transition amplitude of a pion to the vacuum by a nonlocal gauge invariant operator of the axial current .
this amplitude can be expressed in terms of qcd light - cone wave functions @xcite .
assuming that the lfqm can give a good decription for this type of amplitude ( except the region near the end points @xmath1 , as discussed above ) we express the qcd light - cone wave functions of a pion as an integral over the lfqm wave function . using the published parametrizations @xcite for the latter ,
we show that the twist-2 pion wave function is very close to its asymptotic form .
this observation confirms the result obtained in @xcite based on constraints for the light - cone wave function determined from qcd sum rules@xcite for @xmath13 coupling constant @xcite and from the qcd sum rule for the pion structure function @xcite .
the constraint obtained in @xcite also indicates that the the light - cone wave function is close to its asymptotic form . in the lfqm
, we also evaluate the two - particle twist-4 light - cone pion wave function .
we find the asymptotic form for the twist-4 light - cone wave function to be in good agreement with the lfqm description .
in our work , we derive a connection between lfqm and qcd light - cone wave functions by equating the gauge invariant matrix element [ 1 ] in the two appproaches , where @xmath14={\cal p}e^{ig\int_x^0a_\mu(y)dy_\mu}$ ] .
note that in lfqm considered here the gauge fields are not taken into account .
nevertheless we can assume that in the lfqm the quark wave fuctions correspond to the fixed - point gauge @xmath15 .
then @xmath14=1 $ ] and the matrix element ( [ 1 ] ) can be evaluated in terms of the constituent quark wave functions . in qcd ,
the twist-2 and twist-4 two - particle wave functions of the pion are defined by the following matrix element : & & < 0||d(0)__5u(x)|^+(p ) > + & = & if_p__0 ^ 1e^-iu(px)[_(u)+x^2g_1(u)+x^2g_2(u ) + o(x^4)]du + & + & f_x__0 ^ 1e^-iu(px)g_2(u)du+o(x^2 ) [ vpp ] where @xmath16 is the twist-2 wave function , where @xmath17 and @xmath18 are the twist-4 pion wave functions , where and @xmath19 .
it is useful to first examine the situation where the quark field @xmath20 is placed at the origin , in which case = ip_f _ , [ vp ] where |d(x_2=0)=e^ip_2^x_2_x_2=0|d(p_2^ ) [ vp1 ] and u(x_1=0)=e^-ip_1^x_1 _ x_1=0u(p_1^ ) .
[ vp2 ] following analysis of ref.@xcite , one can determine from eq.([vp ] ) the relationship of the pion decay constant to the lfqm wave function . substituting eqs.([vp1],[vp2 ] ) into eq.([vp ] ) , introducing the lfqm pion wave function @xmath21 defined in terms of the s - wave orbital wave function @xmath22 , ( p)= e^- , [ pwf ] using the analysis of ref.@xcite , and changing variables to the total and relative momenta , one finds that ( @xmath23 ) f_=_0^ , [ vp4 ] where @xmath24 is the constituent mass of a @xmath0 or @xmath25 quark .
in ref.@xcite it was suggested to use the following set of parameters : @xmath26 and @xmath27 @xcite ; this gives @xmath28 . to draw a correspondence with eq.([vpp ] ) , let us consider the case @xmath29 in eq.([vp ] ) .
it is then clear that the @xmath0-quark wave function gets a phase factor @xmath30 . in terms of light - cone variables , p_1 & = & ( p^+ ,
p_1 ^ -,p _ ) + p^&= & p^0p^3 + p_1^&= & p_1 ^ 0p_1 ^ 3 , [ bb1 ] the phase factor has the following form : e^ip_1x = e^-ip^+x^-+ip_x _ , [ bb3 ] where by definition @xmath31 . in eq.([bb3 ] ) we have dropped the @xmath32 term , consistent with the fact that in the lfqm the @xmath32 dependence is usually not considered @xcite . this is justified as long as @xmath33 since , in the infinite momentum frame ( @xmath34 ) , the dependence on @xmath35 , p_1 ^ -= , [ bb2 ] is suppressed .
however , the suppression is absent in the limit when @xmath36 . in the region of small @xmath37
it is incorrect to neglect the @xmath32 dependence , and one therefore can not apply the lfqm for calculations of matrix elements here .
it is clear that one draws the same conclusion for matrix elements where the dominant contribution comes from the region @xmath38 .
it is nevertheless possible to make a comparison to the light - front wave functions at @xmath39 , being mindful of the concerns in the region of the end points , @xmath40 .
these remarks are useful for understanding of limits of lfqm applicability
. then , from eqs.([vpp],[bb3 ] ) one can find that < 0||d(0)_+ _ 5u(x)|(p ) > + = ip_+ e^-ip_1 x , [ vpg ] where we work in the frame of reference in which @xmath41 . the integral over @xmath42 can be rewriten in the form of the integral over the parameter @xmath37 , and the jacobian of the transformation is j= .
[ jac ] as a result of this transformation , eq.([vpg ] ) becomes : & & _ 0 ^ 1 e^ip _ _ + & = & if_p^+_0 ^ 1e^-iu(p^+x^-)[_(u)-x_^2g_1(u)- x_^2g_2(u)]du .
[ lqf ] here we use the fact that @xmath43 when @xmath39 .
> from eq.([lqf ] ) it follows that @xmath44 .
using eq.([lqf ] ) and expanding this expression up to the terms @xmath45 we obtain the following formulae for the twist-2 and twist-4 light - cone wave functions : & & f__(u ) + & = & _ 0^ [ lqff1 ] and & & f _ + & = & _ ^2 , [ lqff2 ] where the additional factor of @xmath46 in eq.([lqff2 ] ) arises from the expansion of the exponential to second order and subsequent integration over the square of the angle between @xmath47 and @xmath48 .
it was noted above that we can not apply the lfqm to determine matrix elements having their dominant contribution occurring near the end points @xmath49 .
therefore , since @xmath44 , we can not expect a good description of the qcd light - cone wave functions in terms of the lfqm in the region near the end points @xmath1 .
the integrals in eqs.([lqff1],[lqff2 ] ) may be expressed in terms of the incomplete gamma function @xcite @xmath50 .
the result is f__(u)=me^ ( , z_0 ) [ fr1 ] and f_=e^ m^3[(,z_0)-(,z_0 ) ] , [ fr2 ] where @xmath51 .
we compare our results with asymptotic qcd light - cone wave functions : _
( u)&=&6u|u + g_1(u)&=&^2u^2|u^2 + g_2(u)&=&^2|uu(u-|u ) + g_2(u)&=&^2|u^2u^2 , [ asym ] where @xmath52 , and with constraints on the light - cone wave function obtained from the qcd light - cone sum rule for the @xmath13 coupling constant @xcite and from the light - cone qcd sum rule for the pion structure function @xcite .
the qcd sum rule estimate @xcite for @xmath53 yields @xmath54 gev@xmath4 . comparing eq.([asym ] ) to eqs.([fr1],[fr2 ] ) , we find the results shown in fig.1 and fig.2 , respectively .
we see that the agreement for the twist-2 wave function is about 10% and for the twist-4 wave function is 12 % at the peaks , @xmath55 .
the lfqm predicts a broader distribution and is consistent with results of ref.@xcite .
we have obtained a connection between the lfqm wave function of the pion and the corresponding two - particle wave functions in qcd by identifying matrix elements of quark fields , @xmath5 .
this connection has permitted us to check conclusions obtained from the qcd sum rule analysis for the pion against findings in the lfqm .
we also found a simple reason to be wary of the comparison between the lfqm predictions and the qcd light - cone wave functions in the region where @xmath1 : one of the main lfqm results , namely the absence of a sensitivity to the @xmath32 component of the quark momentum , does not permitt us to take seriously the comparisons near the end points @xmath1 .
it was shown that the lfqm description of qcd light - cone wave functions indicates not only that the twist-2 pion light - cone wave function is close to its asymptotic form but also that the form of the twist-4 light - cone wave function @xmath56 is not far from asymptotic .
one of the authors ( v.b . )
thanks i.narodetsky for stimulating discussions .
this research was sponsored in part by the u.s .
department of energy at los alamos national laboratory under contract w-7405-eng-36 and by the intas grant 93 - 0283 and by the crdf grant rp-2 - 132 .
terentev , yad . fiz . * 24 * ( 1976 ) 207 [ sov .
* 24 * ( 1976 ) 106 ] ; v.b .
berestetsky and m.v .
terentev , _ ibid _ * 24 * ( 1976 ) 1044 ; [ * 24 * ( 1976 ) 547 ] * 25 * ( 1977 ) 653 [ * 25 * ( 1977 ) 347 ] .
terentev , yad . fiz .
* 38 * ( 1983 ) 213 [ * 38 * ( 1983 ) 124 ] .
aznaurian , a.s .
bagdasarian , and n.l .
ter - isaakian , phys . lett . *
b112 * 393 ; i.g .
aznauryan and k.a .
oganessyan , phys . lett . * b249 * ( 1990 ) 309 ; p.l .
chung , f. coester , and w.n .
polyzou , phys.lett .
* b205 * ( 1988 ) 545 ; p.l .
chung , f. coester , b.d .
keister , and w.n .
polyzou , phys.rev .
c * 37 * ( 1988 ) 2000 ; f. cardarelli , i.l .
grach , i.m .
narodetskii , g. salme , and s. simula , phys.lett * b359 * ( 1995 ) 1 ; f. cardarelli , i.l .
grach , i.m .
narodetskii , e. pace , g. salme , and s. simula , phys.rev .
* d53 * ( 1996 ) 6682 .
zhitnitsky , in minneapolis 1996 , continuous advances in qcd , 345 - 357 ; e - print archive : hep - ph/9605226 .
n. isgur , in the new aspects of subnuclear physics , edited by a. zichichi ( plenum publishing corporation , new york , 1980 ) ; + j. rosner , in techniques and concepts of high energy physics , edited by t. ferbel ( plenum publishing corporation , new york , 1981 ) .
brodsky , t. huang and p. lepage , in particles and fields , edited by a.z .
capri and a.n .
kamal ( plenum publishing corporation , new york , 1983 ) .
morgunov , v.i .
shevchenko , and yu.a .
simonov , e - print archive : hep - ph/9704282 .
v.l . chernyak and a.r .
zhitnitsky , phys.rep .
* 112 * ( 1984 ) 173 .
efremov and a.v .
radyushkin , phys.lett .
* b94 * ( 1980 ) 245 .
lepage and s.j .
brodsky , phys .
d * 22 * ( 1980 ) 2157 .
w. jaus , phys .
d * 44 * ( 1991 ) 2851 .
belyaev and m.b .
johnson , la - ur-97 - 1119 , spht t97/032 and e - print archive : hep - ph/9703244 .
shifman , a.i .
vainshtein , and v.i .
zakharov , nucl .
b147 * ( 1979 ) 385 , 448 .
balitsky , v.m .
braun , and a.v .
kolesnichenko , yad.fiz .
* 41 * ( 1985 ) 282 .
braun and i.e. filyanov , z.phys .
* c44 * ( 1989 ) 157 .
belyaev and m.b .
johnson , to be published in phys .
d ; la - ur-97 - 1118 , spht t97/010 and e - print archive : hep - ph/9702207 .
v.m . belyaev and m.b .
johnson , in minneapolis 1996 , continuous advances in qcd , 338 - 344 ; e - print archive : hep - ph/9605279 .
a.v . radyushkin and r. ruskov , nucl.phys . *
b481 * ( 1996 ) 625 . m. abramowitz and i.a .
stegun , _ handbook of mathematical functions _
( dover publications , inc . , new york , 1965 ) .
novikov , m.a .
shifman , a.l .
vainshtein , and v.i zakharov , nucl . phys . *
b237 * ( 1984 ) 525 . * figure captions * + + figure 1 . comparison between the asymptotic twist-2 light - cone wave function @xmath16 ( solid curve ) and the result of the lfqm ( dashed curve ) , given by eq.([fr1 ] ) .
the points at @xmath57 and @xmath55 are constraints determined in refs.@xcite and @xcite , respectively .
+ + figure 2 .
comparison between the asymptotic twist-4 light - cone wave function @xmath56 ( solid curve ) and the result of the lfqm ( dashed curve ) , given by eq.([fr2 ] ) . | we discuss a relation between the light - front quark model and qcd .
we argue that this model can be used for an evaluation of the light - cone wave functions for moderate values of @xmath0 , but that it is inapplicable for this purpose in the region near the end points @xmath1 .
we find additional support for a recent analysis in which it was claimed that the twist - two pion wave function attains its asymptotic form .
the asymptotic twist - four two - particle wave function is also in good agreement with the light - front quark model .
pacs number(s ) : 11.15.tk , 12.38.lg , 12.39.ki , 12.39.-x , 14.40.aq spht - t97/ + july 1997 * pion light - cone wave functions and light - front quark model * v.m .
belyaev@xmath2 _ spht , cea - saclay , 91191 gif - sur - yvette , cedex , france _ mikkel b. johnson + los alamos national laboratory , los alamos , nm 87545 , usa |
the development of the far - off - resonant optical techniques for trapping ultracold atomic gases has opened new directions in the studies of bose - einstein condensates ( becs ) .
atoms can be confined regardless of their hyperfine ( spin ) state , thus avoiding freezing the atom s spin degree of freedom , making it available for the study of interesting spin dynamics @xcite .
one of the major achievements in this direction was the experimental creation of _ spinor _ becs @xcite .
the spinor condensate formed by atoms with spin @xmath1 is described by a @xmath2-component macroscopic ( mean - field ) wave function , which gives rise to various phenomena that are not present in single - component becs , including formation of spin domains @xcite , spin textures @xcite , and multi - component ( vectorial ) solitons of bright @xcite , dark @xcite , gap @xcite and bright - dark @xcite types . generally , the dynamics of the spinor @xmath0 bec is spin - mixing @xcite .
however , there also exist non - mixing , or _
spin - polarized _ states of the system , which are stationary solutions of the corresponding system of gross - pitaevskii ( gp ) equations @xcite .
the stability of such polarized states is different for the two distinct types of the @xmath0 condensates , namely the _ ferromagnetic _
( fm ) , such as in @xmath3rb , and _ polar _ ( alias _ _ anti - ferromagnetic _ _ , afm ) , such as @xmath4na , ones , where the spin - dependent interactions are , respectively , attractive and repulsive .
accordingly , as demonstrated in ref .
@xcite ( see also ref .
@xcite ) , spin - polarized states are modulationally stable / unstable in the afm / fm condensates .
in this work , we focus on afm spinor condensates , and study , in particular , spin - polarized states of the spinor bec of @xmath4na atoms . assuming that this spinor system is confined in a strongly anisotropic trap , we first present the respective system of three coupled quasi one - dimensional ( 1d ) gp equations .
then , employing the so - called single - mode approximation @xcite , we use analytical and numerical methods to find the spin - polarized states of the system and study their stability via the bogoliubov - de gennes ( bdg ) equations ( i.e. , the linearization of the gp equations ) .
the simplest possible form of these states are the stable thomas - fermi ( tf ) configurations ( for each hyperfine component ) .
we also present other spin - polarized states , including ones in which one component maintains a pulse - like shape , inducing a hole in the other components , and structures with holes in all the three components .
all these states feature instability regions too , but with values of the normalized instability growth rate up to @xmath5 .
development of the oscillatory instabilities is examined by means of direct simulations .
it is found that the structures with a hole in one component get weakly deformed , while the states with holes in all the three components suffer stronger deformations .
if the three components are initially spatially separated , being confined in different harmonic traps , but are put in a single trap afterwards , spin domain - wall ( dw ) patterns are formed .
a family of the dw solutions exists and is fully stable if , for a fixed value of the trap s strength , the chemical potential ( or the number of atoms ) exceeds a certain critical value .
the paper is organized as follows : section ii presents the model .
sections iii is dealing with tf states . in sections
iv and v we examine structures with one and multiple holes , respectively ( including their stability ) .
finally , conclusions are presented in section vi .
at sufficiently low temperatures , and in the framework of the mean - field approach , a spinor bec with @xmath0 is described by a vectorial order parameter , @xmath6^{t}$ ] , with the different fields corresponding to three values of the vertical component of the spin , @xmath7 . assuming that this condensate is loaded into a strongly anisotropic trap , with holding frequencies @xmath8 , we assume , as usual , that the wave functions are separable , @xmath9 , where the transverse components @xmath10 represent the ground state of the respective harmonic oscillator . then , following the standard approach @xcite of averaging the coupled 3d gp equations for the three components in the transverse plane @xmath11 , leads to the system of coupled 1d equations for the longitudinal components of the wave functions ( see refs .
@xcite ) : @xmath12where star denotes complex conjugate , and @xmath13 is the spin - independent part of the hamiltonian , with @xmath14 being the total density ( @xmath15 is the atomic mass ) .
the nonlinearity coefficients have an effectively 1d form , namely @xmath16 and @xmath17 , where @xmath18 is the transverse harmonic - oscillator length which determines the size of the transverse ground state . coupling constants @xmath19 and @xmath20 account for , respectively , the mean - field spin - independent and spin - dependent binary interactions between identical spin-@xmath21 bosons , @xmath22where @xmath23 and @xmath24 are the @xmath25-wave scattering lengths in combined symmetric collision channels of total spin @xmath26 .
the spinor @xmath0 condensate with @xmath27 and @xmath28 is , respectively , of the fm and afm types , as , respectively , in @xmath3rb and @xmath4na @xcite .
measuring time , length and density in units of @xmath29 , @xmath30 and @xmath31 ( here , @xmath31 is the peak density ) , we cast eqs .
( [ mgp1])-([mgp2 ] ) in the dimensionless form , @xmath32 where @xmath33 , the normalized trap s strength is @xmath34 and the parameter @xmath35 is given by @xmath36 ( @xmath37 and @xmath38 correspond , respectively , to the fm and afm spinor condensate ) . for spin-@xmath21 @xmath3rb and @xmath4na atoms ,
this parameter is @xmath39 @xcite and @xmath40 @xcite , respectively .
thus , in these physically relevant cases , @xmath41 is a fixed small parameter of the system .
_ spin - polarized states _ of the system , characterized by a constant population of each spin component , can be found upon substitution @xmath42where @xmath43 and @xmath44 are densities and phases of the components , and @xmath45 are their chemical potentials . substituting this in eqs .
( dvgp1 ) and ( [ dvgp2 ] ) , it is readily found that conditions for the existence of spin - polarized states are @xmath46below , we assume that the chemical potentials of the components are equal , @xmath47 . for analysis of the stability of a stationary spin - polarized state ,
@xmath48^{t}$ ] , we will perform linearization around the unperturbed spin - polarized state , assuming a perturbed solution , @xmath49 e^{-i\mu t } , \label{stansatz}\]]where @xmath50 and @xmath51 represent infinitesimal perturbations with eigenvalues @xmath52 .
then , the solution of the ensuing linear - stability problem ( i.e. , the bdg equations ) for @xmath53 and associated eigenfunctions @xmath54 provides complete information about the stability of the underlying stationary state , @xmath55 .
whenever it is unstable , we will also examine its evolution through direct simulations of gp equations ( [ dvgp1 ] ) and ( [ dvgp2 ] ) , using a finite difference scheme in space and a fourth - order runge - kutta time integrator .
typically , in the simulations the unstable state is initially perturbed by a uniformly distributed random perturbation of relative amplitude @xmath56 . to estimate relevant physical parameters
, we assume the spinor condensate of @xmath4na atoms with peak 1d density @xmath57 m@xmath58 , confined in the harmonic trap with frequencies @xmath59hz and @xmath60hz . in this case , the normalized trap strength is @xmath61 ( this value is kept fixed throughout this work ) , while the number of atoms , @xmath62 , depends on chemical potential @xmath63 and the particular form of spin - polarized states .
typically , we take @xmath63 in interval @xmath64 , which corresponds to values of @xmath62 in the range of @xmath65 atoms . for instance , at @xmath66 the number of atoms is @xmath67 ; in this case , the normalized time and space units correspond , respectively , to @xmath68 ms and @xmath69 m .
the simplest spin - polarized states can be found in the framework of the _ single - mode approximation _ @xcite . in anticipation ,
suggested by the approximation that components @xmath70 may be close to eigenmodes of a single effective potential , induced by the combination of the trap and nonlinearity , we introduce the ansatz @xmath71 , where coefficients @xmath72 are the populations of each spin component in the steady state , related by the normalization condition , @xmath73 . then , eqs .
( [ dvgp1])-([dvgp2 ] ) lead to the following system : @xmath74 \sqrt{n}=0,\quad \label{in1 } \\[2.0ex ] \left [ \hat{l}+\delta \left ( 1-q_{0}+2s\sqrt{q_{-1}q_{+1}}\right ) n\right ] \sqrt{n}=0 , \quad \label{in2}\end{aligned}\ ] ] where @xmath75 and @xmath76 for @xmath77 or @xmath78 , respectively .
it is readily observed that the three equations ( [ in1 ] ) , ( [ in2 ] ) are tantamount to a single one , viz .
@xmath79supplemented by relations @xmath80 & & p\equiv 1,\,\,\,\ q_{-1}=q_{+1}\,\,\,\ \mathrm{for}\,\,\,\ \delta \theta = \pi .
\label{con2}\end{aligned}\ ] ] in uniform space ( @xmath81 ) , eq .
( [ gs ] ) leads to the constant density , @xmath82 . as shown in refs .
@xcite , such constant solutions to eqs .
( [ dvgp1])-([dvgp2 ] ) are modulationally stable only in the afm phase , with @xmath38 .
below , we will only consider the case of the afm condensate ( in particular , bosonic sodium atoms with spin @xmath21 and @xmath83 ) .
in the presence of a sufficiently weak trap , eq .
( [ gs ] ) can be solved approximately employing the tf approximation @xcite : neglecting the kinetic - energy term ( @xmath84 ) in eq .
( [ gs ] ) , we find density profiles of the three spin components : in the region where @xmath85 , @xmath86and @xmath87 elsewhere . obviously , all three components of the tf solution have the same spatial width , or thomas - fermi radius , @xmath88 . in numerical simulations , we used a fixed - point algorithm ( newton - raphson method ) to find exact spin - polarized solutions to eqs .
( [ dvgp1])-([dvgp2 ] ) , with profiles close to those produced by the tf approximation , as given by eq .
( [ tf ] ) .
in particular , we used , as an initial guess , three identical profiles of the form @xmath89 with @xmath90 and @xmath77 or @xmath91 .
then , keeping the trap s strength , @xmath92 , fixed , we varied the chemical potential @xmath63 , and the numerical solution converged to stable spin - polarized states , which were indeed close to the approximate one given by eq .
( [ tf ] ) .
two typical examples are shown in the top panels of fig .
[ fig1 ] for both cases , @xmath77 ( left panel ) and @xmath78 ( right panel ) , with @xmath61 and @xmath93 .
the numerically determined states are very close to their tf - predicted counterparts , with @xmath94 and @xmath95 ( @xmath77 , left panel ) and @xmath96 and @xmath97 ( @xmath78 , right panel ) .
( left panel ) and @xmath98 ( right panel ) , both obtained for @xmath61 and @xmath99 .
the wavefunctions @xmath100 are identical and are depicted by the solid line , while the wavefunction @xmath101 is depicted by the dashed line . in the left panel , @xmath94 and @xmath95 , while in the right panel @xmath96 and @xmath102 . , title="fig:",width=155,height=124 ] ( left panel ) and @xmath98 ( right panel ) , both obtained for @xmath61 and @xmath99 .
the wavefunctions @xmath100 are identical and are depicted by the solid line , while the wavefunction @xmath101 is depicted by the dashed line . in the left panel , @xmath94 and @xmath95 , while in the right panel @xmath96 and @xmath102 .
, title="fig:",width=147,height=128 ]
apart from the tf spin - polarized states , there exist other ones , which feature holes in some of the components .
an example of such states is shown in fig .
[ fig2 ] . as seen in the two top panels of this figure ( both a stable and an unstable state
are shown see below ) , the @xmath103 component is concentrated in the form of a pulse located at the trap s center , while the @xmath104 components feature a large hole at the same spot .
this arrangement is explained by the fact that the interaction between components is repulsive , hence a peak ( hole ) in @xmath103 ( @xmath104 ) induces a hole ( peak ) in @xmath104 ( @xmath103 ) .
the norm @xmath62 of each component is shown , as a function of chemical potential @xmath63 , in the third - row panel of fig .
the set of the linear stability eigenvalues for this state is shown in the second - row panels of fig .
[ fig2 ] , for two different values of @xmath63 : the state with @xmath105 ( top - right ) is stable , as all the eigenvalues are imaginary , while the state with @xmath93 ( top - left panel ) is unstable . in fact , all such unstable states are destabilized by a hamiltonian hopf bifurcation , which leads to a quartet of eigenvalues with nonzero real parts .
the instability interval is @xmath106 , with the maximum instability growth rate max@xmath107 , found at @xmath108 ( see bottom panel of fig .
[ fig2 ] ) .
components , and a pulse - like @xmath101 component , for @xmath61 and @xmath109 .
the two top left and right panels show unstable and stable states , with @xmath110 and @xmath111 , respectively ; solid and dashed lines depict the components @xmath112 and @xmath101 .
the two panels in the second row display the spectral planes @xmath113 of the ( in)stability eigenvalues for the same states . note that the instability ( at @xmath110 ) is of the oscillatory type , accounted for by a quartet of eigenvalues with nonzero real parts .
the panel in the third row shows the normalized number of atoms ( norm ) , @xmath62 , of each component as a function of chemical potential @xmath114 ; solid and dashed lines show the norms of components @xmath100 and @xmath115 , respectively . the bottom panel shows @xmath116 as a function of @xmath114 , which reveals the instability window at @xmath117 , with a maximum instability growth rate @xmath118 at @xmath119 .
the latter value corresponds to the unstable state shown in the top - left panel.,title="fig:",width=302 ] components , and a pulse - like @xmath101 component , for @xmath61 and @xmath109 .
the two top left and right panels show unstable and stable states , with @xmath110 and @xmath111 , respectively ; solid and dashed lines depict the components @xmath112 and @xmath101 .
the two panels in the second row display the spectral planes @xmath113 of the ( in)stability eigenvalues for the same states . note that the instability ( at @xmath110 ) is of the oscillatory type , accounted for by a quartet of eigenvalues with nonzero real parts .
the panel in the third row shows the normalized number of atoms ( norm ) , @xmath62 , of each component as a function of chemical potential @xmath114 ; solid and dashed lines show the norms of components @xmath100 and @xmath115 , respectively . the bottom panel shows @xmath116 as a function of @xmath114 , which reveals the instability window at @xmath117 , with a maximum instability growth rate @xmath118 at @xmath119 .
the latter value corresponds to the unstable state shown in the top - left panel.,title="fig:",width=302 ] a similar state with the roles of @xmath104 and @xmath103 exchanged , i.e. , the @xmath103 component featuring the hole , and @xmath104 ones concentrated in narrow pulses , were also found .
moreover , such states were found with either @xmath120 or @xmath121 having the opposite sign ( i.e. , for @xmath122 ) .
the results are not shown here , as the ( in)stability of these states is qualitatively the same as in the above case .
the evolution of unstable states is exemplified in fig .
[ fig4 ] , which displays results of direct simulations of eqs .
( [ dvgp1 ] ) and ( [ dvgp2 ] ) for the unstable state with @xmath77 and @xmath93 , that was presented in fig .
[ fig2 ] ( top - left panel ) . in fig .
[ fig4 ] , contour plots of densities of the components of the solution are displayed as a function of time ( the densities of the @xmath123 and @xmath124 components are identical ) . it is clearly observed that the predicted oscillatory instability sets in at a very large time ( @xmath125 , which corresponds to @xmath126 seconds in physical units ) ; this is a consequence of the extremely small growth rate of the instability .
eventually , the system settles down to a steady state , which has many similarities with the initial one . in particular , as seen in the top panels of fig .
[ fig4 ] , after @xmath127 the pulse in the @xmath103 component broadens and its amplitude is accordingly decreased , while the hole in the @xmath104 components becomes shallower .
it is also noted that the three components develop a similar structure in their tails , as shown in the bottom panels of fig .
[ fig4 ] .
subject to the oscillatory instability .
the left and right panels display , respectively , identical densities of @xmath100 and that of @xmath101 .
because of the extremely small growth rate of the instability , it manifests itself only for @xmath128 .
bottom panels : the respective wavefunction profiles at @xmath129 ( solid lines ) and @xmath130 ( dashed lines ) ; as above , the left and right panels show , respectively , the wavefunctions @xmath100 and @xmath101 . ,
title="fig:",width=154 ] subject to the oscillatory instability .
the left and right panels display , respectively , identical densities of @xmath100 and that of @xmath101 . because of the extremely small growth rate of the instability , it manifests itself only for @xmath128 .
bottom panels : the respective wavefunction profiles at @xmath129 ( solid lines ) and @xmath130 ( dashed lines ) ; as above , the left and right panels show , respectively , the wavefunctions @xmath100 and @xmath101 . ,
title="fig:",width=154,height=128 ] subject to the oscillatory instability .
the left and right panels display , respectively , identical densities of @xmath100 and that of @xmath101 . because of the extremely small growth rate of the instability , it manifests itself only for @xmath128 .
bottom panels : the respective wavefunction profiles at @xmath129 ( solid lines ) and @xmath130 ( dashed lines ) ; as above , the left and right panels show , respectively , the wavefunctions @xmath100 and @xmath101 . ,
title="fig:",width=154 ] subject to the oscillatory instability .
the left and right panels display , respectively , identical densities of @xmath100 and that of @xmath101 . because of the extremely small growth rate of the instability , it manifests itself only for @xmath128 .
bottom panels : the respective wavefunction profiles at @xmath129 ( solid lines ) and @xmath130 ( dashed lines ) ; as above , the left and right panels show , respectively , the wavefunctions @xmath100 and @xmath101 . , title="fig:",width=154 ] apart from the states considered above , it is also possible to find spin - polarized states which feature , e.g. , one hole in each of the @xmath131 components , and two holes in @xmath132 .
examples of such a state are shown in the top panels of fig .
[ fig5 ] ( the left one , for @xmath93 , is stable , while the right one , for @xmath105 is unstable see below ) . as seen in this figure
, one may consider @xmath124 and @xmath123 as built of two overlapping pulses , which induce two holes in the @xmath103 component due to the repulsive interatomic interactions .
results of the stability analysis for these states are shown in fig .
[ fig5 ] . in this case ,
two quartets of eigenvalues with nonzero real parts are found in the spectral plane ( see the second row right panel in fig .
[ fig5 ] ) in the interval @xmath133 ( the bottom panel in fig .
[ fig5 ] ) .
the respective largest instability growth rate is @xmath134 for @xmath135 , i.e. , of the same order of magnitude as in the previous case .
the development of the instability was studied , as above , in direct simulations , starting with initial conditions in the form of a perturbed solution pertaining to @xmath105 .
the result is shown in fig .
[ fig6 ] , in terms of the evolution of identical densities of the @xmath104 components , and the density of @xmath103 .
again , the instability manifests itself at large times ( @xmath136 , which corresponds to @xmath137 seconds in physical units ) , but in this case the final result is a strong oscillatory deformation of the three spin components ( after @xmath138 ) , contrary to what is the case in fig .
[ fig4 ] ( the establishment of a new spin - separated state ) .
but for a state with one hole in each of the @xmath100 components and two holes in the @xmath101 component . in this case ,
the instability is manifested by two eigenvalue quartets the largest of which has a nonzero real part in the interval @xmath139 ; the maximum instability growth rate is @xmath140 for @xmath141 .
the unstable state shown in the top - right panel corresponds to @xmath142 . ,
title="fig:",width=302 ] but for a state with one hole in each of the @xmath100 components and two holes in the @xmath101 component . in this case , the instability is manifested by two eigenvalue quartets the largest of which has a nonzero real part in the interval @xmath139 ; the maximum instability growth rate is @xmath140 for @xmath141 . the unstable state shown in the top - right panel corresponds to @xmath142 .
, title="fig:",width=302 ] , for the unstable state shown in the top - right panel of fig . [ fig5 ] ( pertaining to @xmath142 ) . the instability manifests itself at large times ( @xmath143 ) and results in strong oscillatory deformation of the clouds .
, title="fig:",width=154,height=128 ] , for the unstable state shown in the top - right panel of fig . [ fig5 ] ( pertaining to @xmath142 ) . the instability manifests itself at large times ( @xmath143 ) and results in strong oscillatory deformation of the clouds .
, title="fig:",width=154,height=128 ] , for the unstable state shown in the top - right panel of fig . [ fig5 ] ( pertaining to @xmath142 ) . the instability manifests itself at large times ( @xmath143 ) and results in strong oscillatory deformation of the clouds .
, title="fig:",width=154 ] , for the unstable state shown in the top - right panel of fig . [ fig5 ] ( pertaining to @xmath142 ) . the instability manifests itself at large times ( @xmath143 ) and results in strong oscillatory deformation of the clouds .
, title="fig:",width=154 ] similar states with one hole in @xmath103 and two holes in each of the @xmath144 components , as well their counterparts with @xmath145 , have also been found .
they are not shown here ; as in the previous case , the ( in)stability of these additional states is similar to that reported in fig .
[ fig5 ] .
, @xmath101 and @xmath146 are confined in harmonic traps of strength @xmath61 located , respectively , at @xmath147 , @xmath148 ( dashed lines ) and @xmath149 ( solid line ) .
the traps at @xmath147 and @xmath150 are turned off and this spinor configuration , confined solely in the trap located at @xmath149 , is introduced in the fixed - point algorithm .
, width=226 ]
in the above sections , we reported the spin - polarized states in which all three spin components were spatially overlapping , since they were confined to the same potential trap .
however , it is also possible to use three different traps , each confining a different component , to initially separate them , and then allow the system to evolve in the presence of one of these traps ( i.e. , turning off the other two ) . in this section , we present spin - polarized states , including dw structures , obtained in this way .
first , we describe the initialization of the system .
we assume that the three tf - shaped components are initially loaded into three different traps , @xmath151 , of the same strength , @xmath92 , centered at different positions : @xmath152we choose @xmath153 ( i.e. , @xmath154 for @xmath61 ) , which implies initially overlapping tf configurations , see fig . [ fig7 ] . after constructing this state , we turn off the traps @xmath155 and @xmath156 , retaining only the rightmost one , @xmath157 , which now acts on all the three components , and feed the initial configuration state into the fixed - point algorithm , to find new spin - polarized states .
other possibilities , such as turning off potentials @xmath158 and keeping @xmath159 , arranging the three components in a different way , etc . , eventually lead to retrieving the spin - polarized states presented in the previous sections , while the approach outlined above [ keeping @xmath157 and switching @xmath155 and @xmath156 off ] generates new dw patterns , which are displayed in fig .
[ fig8 ] , and could not be obtained otherwise .
( the asymmetry of the procedure is instrumental in generating the new states ) . and
@xmath160 are identical ) in a stationary state found from the initial configuration prepared as shown in fig .
the resulting spin - polarized state has the form of a domain - wall like structure between the @xmath101 and @xmath100 components .
the parameters are @xmath61 and @xmath93 .
top - right panel shows the wavefunctions of the domain - wall state found at @xmath161 .
bottom panel : the norm of each component in the above - mentioned domain - wall structure versus the chemical potential ( the dependences for @xmath109 or @xmath162 are identical ) .
, title="fig:",width=154,height=147 ] and @xmath160 are identical ) in a stationary state found from the initial configuration prepared as shown in fig .
the resulting spin - polarized state has the form of a domain - wall like structure between the @xmath101 and @xmath100 components .
the parameters are @xmath61 and @xmath93 .
top - right panel shows the wavefunctions of the domain - wall state found at @xmath161 .
bottom panel : the norm of each component in the above - mentioned domain - wall structure versus the chemical potential ( the dependences for @xmath109 or @xmath162 are identical ) .
, title="fig:",width=154,height=138 ] and @xmath160 are identical ) in a stationary state found from the initial configuration prepared as shown in fig .
the resulting spin - polarized state has the form of a domain - wall like structure between the @xmath101 and @xmath100 components .
the parameters are @xmath61 and @xmath93 .
top - right panel shows the wavefunctions of the domain - wall state found at @xmath161 .
bottom panel : the norm of each component in the above - mentioned domain - wall structure versus the chemical potential ( the dependences for @xmath109 or @xmath162 are identical ) .
, title="fig:",width=302 ] the most interesting spin - polarized dw states found following this procedure correspond to values of the chemical potential @xmath163 ( or norm @xmath164 for @xmath61 ) ; for smaller values of @xmath165 we typically find structures of the thomas - fermi type .
two examples , one for @xmath166 , and another corresponding to @xmath167 , are shown in fig .
[ fig8 ] . in the former one , the @xmath168 component ( which has the larger norm ) is centered to the right of the midpoint of the remaining trap ( @xmath169 ) , while the identical @xmath104 components are pushed to the left , due to the repulsion from @xmath103 , with a dw created between @xmath104 and @xmath103 .
figure [ fig8 ] displays an example of such a dw structure for @xmath93 and @xmath61 .
the state found at the above - mentioned value , @xmath167 , is shown in the top - right panel of fig .
[ fig8 ] , for @xmath77 . in this state , the shape of the @xmath103 component is similar to that considered above , while the @xmath121 components are _ not _ identical , in contrast to the above example . in this case
the @xmath103 and @xmath104 components overlap over a wider spatial region . note that @xmath170 changes its sign at @xmath171 , featuring a structure resembling the waveform of a dark soliton embedded in a bright one @xcite .
the stability of the dw states was also investigated by means of the bdg equations .
it was concluded that there are no eigenvalues with a real part in the interval @xmath172 , or , equivalently , @xmath173 for @xmath61 ( not shown here in detail ) .
thus , the dws are stable in this region .
verification of the stability , performed by direct simulations of eqs .
( [ dvgp1 ] ) and ( [ dvgp2 ] ) , is illustrated in figs .
[ fig9 ] and [ fig10 ] , for @xmath93 and @xmath174 respectively .
it is obvious that these states are indeed stable , for very large times exceeding @xmath130 ( i.e. , @xmath175 seconds in physical units ) . , to which a random perturbation was added . shown in the left and right panels are spatiotemporal contour plots of the densities of components @xmath100 ( identical ones ) and @xmath101 , respectively.,title="fig:",width=151,height=113 ] , to which a random perturbation was added .
shown in the left and right panels are spatiotemporal contour plots of the densities of components @xmath100 ( identical ones ) and @xmath101 , respectively.,title="fig:",width=151,height=113 ] but for the state shown in the top - right panel of fig .
the left , middle and right panels show , respectively , the densities of @xmath160 , @xmath115 and @xmath146 .
noteworthy is a stationary dark - soliton - like structure , located at @xmath171 in the @xmath176 component . ,
title="fig:",width=94,height=75 ] but for the state shown in the top - right panel of fig .
[ fig7 ] . the left ,
middle and right panels show , respectively , the densities of @xmath160 , @xmath115 and @xmath146 .
noteworthy is a stationary dark - soliton - like structure , located at @xmath171 in the @xmath176 component . ,
title="fig:",width=94,height=75 ] but for the state shown in the top - right panel of fig .
[ fig7 ] . the left ,
middle and right panels show , respectively , the densities of @xmath160 , @xmath115 and @xmath146 .
noteworthy is a stationary dark - soliton - like structure , located at @xmath171 in the @xmath176 component .
, title="fig:",width=94,height=75 ]
in this work , we have studied spin - polarized states in anti - ferromagnetic spinor ( @xmath0 ) bose - einstein condensates . in particular
, our analysis applies to a quasi-1d spinor condensate of sodium atoms .
the considerations were based on analytical calculations and numerical computations of the coupled gross - pitaevskii equations for this setting .
assuming that all three hyperfine ( spin ) components are confined in the same harmonic trap , we have found various types of spin - polarized states and examined their stability .
the first family consists of thomas - fermi configurations , considered analytically in the framework of the single - mode approximation ( which assumes the similarity of the spatial profiles of all the components ) . within their existence region , these states were found to be stable .
also identified were more complex patterns , which include states composed of pulse - like structures in one component , which induce holes in the other components , and states with holes in all three components .
these states feature windows of weak instability .
the development of the instability was investigated by means of direct numerical simulations , which demonstrate that it manifests itself at very long times , and results in a weak deformation of the states with a single hole in some of the components , and a stronger one in the states with holes in all components .
fully stable families of spin - polarized states develop from initial configurations which are initially separated ( by means of three different traps ) components .
these states form domain - wall structures between the components , at values of the chemical potential above a critical value . at the critical value ( corresponding to a certain norm , for a fixed trap s strength )
, we have found another spin - polarized state ( including a dark - soliton element ) , in which all the components partly overlap .
it would clearly be interesting to investigate the existence and stability of higher - dimensional counterparts of the 1d spin - polarized states found in this work .
another relevant question for further analysis is whether such spinor condensates support stable topological objects , such as dark solitons or vortices .
work in these directions is in progress .
the work of h.e.n . and d.j.f .
was partially supported from the special research account of university of athens .
acknowledges partial support from ec grants pythagoras i. p.g.k . acknowledges support from nsf - career , nsf - dms-0505663 and nsf - dms-0619492 , as well as the warm hospitality of msri during the initial stages of this work . the work of b.a.m .
was supported , in a part , by the israel science foundation through the center - of - excellence grant no .
8006/03 , and the german - israel foundation through grant no . 149/2006 .
r.c.g . acknowledges support from nsf - dms-0505663 .
work at los alamos national laboratory is supported by the usdoe . | we study spin - polarized states and their stability in anti - ferromagnetic states of spinor ( @xmath0 ) quasi - one - dimensional bose - einstein condensates . using analytical approximations and numerical methods ,
we find various types of polarized states , including : patterns of the thomas - fermi type ; structures with a pulse - shape in one component inducing a hole in the other components ; states with holes in all three components ; and domain walls .
a bogoliubov - de gennes analysis reveals that families of these states contain intervals of a weak oscillatory instability , except for the domain walls , which are always stable .
the development of the instabilities is examined by means of direct numerical simulations . |
wireless networks , the use of relay is attracting increasing attention [ 1 , 2 ] because of its many advantages . among the relay channels , two - way relay channel ( twrc , as shown in fig .
1 ) is a especially interested due to the almost double spectral efficiency with the physical layer network coding ( pnc ) [ 3 ] transmission scheme .
it was further proved in [ 4 , 5 ] that pnc can approach the capacity of twrc in high snr region . another spectral efficiency boosting technique is multiple input and multiple output ( mimo ) , which has been widely used in wireless systems .
therefore , it is of great interest to combine pnc and mimo to further improve the wireless spectral efficiency .
a straightforward way is to divide the mimo transmission into parallel siso streams by precoding , so that pnc can be implemented on each steam [ 6 ] .
however , the precoding requires not only transmitter side channel state information ( csit ) but also strict time and carrier phase synchronization between the two end nodes .
mimo nc scheme [ 7 ] is more practical , where only receiver side csi ( csir ) is needed . in mimo nc , the relay node detects each end node s packet with traditional mimo detection and then combines them with network coding .
these schemes failed to exploit the fact that the relay does not need each end node s individual information .
hence , the performance is limited by over - detection . in [ 8 ] ,
we have proposed a novel mimo pnc scheme based on linear detection , which will be referred as linear mimo pnc in this paper .
linear mimo pnc tries to detect the summation and the difference of the two end node s packets before transforming them to the network coding form .
with similar complexity and csir requirement , it significantly outperforms mimo nc .
however , due to the performance limit of linear detection , linear mimo pnc s performance is poor under bad channel conditions .
besides the linear detection , another popular mimo detection method is vblast ( vertical bell laboratories layered space - time ) which can achieve much better performance with an acceptable increase in complexity .
the vblast architecture was first proposed in [ 9 ] where a code block is de - multiplexed into different layers and each is transmitted through a particular antenna . at the receiver , these layers are successively detected , where the detected interference are canceled and the unknown interferences are nulled by linearly weighting the residual signal vector with a zf ( zero - forcing ) null vector ( zf vblast ) .
a low complexity zf vblast scheme is proposed in [ 10 ] , where the channel matrix is rewritten in terms of the qr decomposition as @xmath0 .
the inverse of unitary matrix @xmath1 was then multiplied to the received signal before estimating the transmit information . in order to find the optimal detection order , [ 10 ] further proposed the sorted qr decomposition algorithm , zf - sqrd . in this paper
, we combine the basic idea of pnc and the qr vblast mimo detection scheme and propose vblast pnc .
our scheme only requires receiver side channel state information and symbol level synchronization between the end nodes , as in general virtual mimo system .
the basic idea of vblast pnc in a 2-by-2 mimo system is as follows .
with qr decomposition , the relay first detects the second layer ( one end node ) signal . rather than canceling all the component of the second layer signal from the first layer as in traditional vblast detection
, we only subtract a part of the second layer information and directly map the residual signal ( including both second and first layer information ) to the network coding form . with such partial interference cancellation , the error propagated from the incorrectly detected second layer is significantly decreased .
thus , the system performance is improved .
we then extend our vblast pnc to a detection scheme with an optimal order as in zf - sqrd , and even better performance is achieved .
numerical simulation is done to compare the performance of vblast pnc with linear mimo pnc and mimo nc schemes .
the results show that vblast pnc can achieve much better ber performance than linear mimo pnc and vblast mimo nc .
this paper is mainly based on the twrc in fig .
1 , where the relay is equipped with 2 antennas and the each end node is equipped with single antenna .
the transmission consists of two phases . in the uplink phase , both
the end nodes transmit their packets to the relay node simultaneously using qpsk modulation .
we assume that the two packets arrive at the relay node in a symbol level synchronization . in that way , the superimposed signal received by the relay is : @xmath2 where @xmath3 denotes the received signal at the @xmath4-th antenna of the relay node ; @xmath5 is zero mean complex gaussian random variable , which denotes the channel coefficient from the end node @xmath6 to the @xmath4-th antenna of the relay node ; @xmath7is the transmitted signal with qpsk modulation of the end node@xmath8 , and @xmath9 demotes the complex gaussian noise with zero mean and @xmath10 variance for each dimension . in this phase ,
the full channel information is available at the relay node .
rewriting the received signal in the vector form as @xmath11 then , the relay node tries to extract some useful information from @xmath12 and transforming it to the network coded form @xmath13 .
the detailed processing will be illustrated later , which is also the focus of our paper . in the downlink phase ,
the relay broadcasts the network coded packet to both end nodes . after receiving the packet from the relay ,
the end nodes extract their target packets with the help of their own information .
now , we present an example to illustrate the basic idea of vblast pnc and its superiority over vblast nc , with a simple channel realization . in the first phase , the transmission in ( 2 )
can be regarded as a @xmath14 mimo system ( two transmit antennas and two received antennas ) .
the relay node s goal is to acquire an estimate of @xmath13 . in the traditional mimo
nc scheme , the relay node decodes @xmath15 and @xmath16 explicitly before network encoding them .
nevertheless , this scheme is suboptimal .
consider an ill - conditioned channel matrix @xmath17 @xmath18\ ] ] where @xmath19 is a small quantity .
then eq.(1 ) can be rewritten as @xmath20 with qr zf vblast detection , @xmath16 is first detected as @xmath21 according to the processing of vblast , we cancel the interference @xmath16 in @xmath22 and obtain the estimate of @xmath15 as @xmath23 from ( [ eq5 ] ) and ( [ eq6 ] ) , we can find that the snr of @xmath15 tends to zero when @xmath19 is very small .
it means that the vblast nc scheme ca nt accurately estimate the target signal @xmath13 in this case .
based on the basic idea of pnc , the relay can estimate the target signal @xmath13 from @xmath24 .
therefore , we can first estimate @xmath24 rather than estimate individual information of @xmath25 , from @xmath12 .
the particular processing is as follows . after estimating @xmath16
, we subtract @xmath26 rather than @xmath27 from @xmath28 .
then , we can obtain @xmath29 with the pnc mapping [ 3 ] , we can directly map @xmath30 in ( 7 ) to the target information @xmath13 .
since the noise in ( 7 ) is small ( independent of @xmath31 , the performance of this scheme is much better than the mimo nc scheme .
this example indicates that mimo pnc may significantly outperform the traditional mimo nc . in the following sections , we introduce the proposed vblast pnc scheme in detail .
in this section , we first briefly review the mimo nc scheme based on qr vblast detection algorithm [ 9 , 10 ] for a comparison . after that
, we elaborate the proposed vblast pnc . consider the 2-by-2 mimo system in ( 2 ) .
the channel matrix can be decomposed with qr decomposition so that @xmath32__qr _ _ , where the matrix @xmath1 is a unitary matrix ( orthogonal columns with unit norm ) and @xmath33 is an upper triangular matrix . multiplying the received signal @xmath12 by @xmath34 ,
we can obtain the calculated signal @xmath35 where @xmath36 denotes the matrix conjugate transpose and the new noise vector @xmath37 has the same distribution as @xmath38 .
the scalar form of @xmath39 is @xmath40 owing to the upper triangular characteristic of @xmath33 , we can easily estimate the second layer signal ( @xmath16 here ) as : @xmath41 since the qpsk modulation is adopted , the above signal can be demodulated with hard decision as @xmath42 note that the hard decision in ( [ eq11 ] ) is performed for the real part and the imaginary part respectively . according to the basic idea of vblast , after the second layer signal @xmath16 is detected , we can then detect the first layer signal ( @xmath15 here ) by canceling the interference of @xmath43as : @xmath44 after obtaining the individual decisions of @xmath15 and @xmath16 , a straightforward way to calculating @xmath13 is to combine the estimates of @xmath15 and @xmath16 ( obtained in ( 11 ) and ( 12 ) respectively ) : @xmath45 hereafter , we refer to this scheme as vblast nc .
vblast nc detects the signals separately with vblast algorithm , and then encodes them into the network - coded form.however , it may perform poorly as shown in our illustrating example . as in the vblast processing above , exchange of the detection order between x1 and x2 ( first detect @xmath15 and
then cancel it before detecting @xmath46 is also workable . as shown in [ 10 ] , the detection sequence is crucial to the performance of vblast because of the risk of error propagation . to obtain a better performance , we can permute the columns of the channel matrix @xmath17 before the qr decomposition . by carefully selecting the permutation pattern ,
the post - permuted @xmath33 has a larger @xmath47 . as a result ,
the snr of the second layer is increased and the ber is decreased , and the error propagation effect is also decreased .
we can also adopt this sorted algorithm [ 10 ] in our vblast pnc , as elaborated in the following part . in vblast ,
the diversity order of the first layer should be larger than the second layer in theory under the assumption of clean interference cancellation . in practice
, however , the diversity order of both layers is the same [ 11 ] .
the main reason is the error propagated from the erroneously detected second layer signal . to mitigate the error propagation and better the performance , we propose vblast pnc scheme which only cancels a part of the detected second layer information .
after the cancellation , we require the remaining signal to be in the form of @xmath48 where @xmath49 is an integer .
we can then directly map this signal to the target signal @xmath50 by applying pnc mapping , without explicitly detecting @xmath15 .
since the pnc mapping has similar performances as the ordinary point - to - point transmission , our vblast pnc could achieve a better performance by mitigating the error propagation effects .
the detail of vblast pnc is as follows .
we rewrite @xmath51 in ( 9 ) as @xmath52 where @xmath49 is an integer to be determined later . in ( [ eq1 ] )
we regard @xmath53__kx__@xmath54 as the signal to be estimated and ( @xmath55__-kr__@xmath56 as interference to be cancelled . in order to decrease the effect of error propagation
, we must minimize ( @xmath55__-kr__@xmath57 .
for example , if @xmath58 , no interference needs to be cancelled and there will never exist error propagation during the detection of the signal in ( [ eq14 ] ) . taking the integer requirement of @xmath49 into account
, we can determine the value of @xmath49 as @xmath59 where @xmath60 means the integer nearest to @xmath61 . in ( 15 )
, @xmath55 is a complex variable and @xmath62 is a real variable , and we only take account the real part to calculate @xmath49 in this paper . after cancelling the interference with the hard estimate of @xmath16 in ( 11 ) ,
we can obtain the estimate of @xmath63 as @xmath64 finally , the relay node estimates @xmath13 only from @xmath63 as long as @xmath49 does not equal to 0 . when k>0 , intuitively , @xmath65 is larger for @xmath66 , while it is smaller for @xmath67 . when k<0 , the situation reverse
. then , for each dimension ( real part or imaginary part ) signal , the corresponding decision rule is @xmath68 where @xmath69 is the decision threshold and its optimal value can be calculated as in [ 3 ] . in high snr , we can simplify the calculation of @xmath69 and set it to @xmath70
. then we have @xmath71 to further improve the performance , we can extend our scheme to the sorted vblast pnc , where the optimal detection order is chosen as the tradition vblast .
in particular , we exchange the columns of @xmath17 to obtain a larger @xmath47 .
then , the vblast algorithm is performed on the new @xmath17 .
in this section , we present the simulation results for the proposed vblast pnc . to compare its performance
, we also show the simulation performance of linear mimo pnc , the performance of linear mimo nc and vblast nc .
the simulation setting is mainly based on the system model in section ii . in particular
, we use qpsk modulation and set the packet length to @xmath72 .
the wireless channels are assumed to be block fading with each entry of the channel matrix to be independently complex gaussian distributed over @xmath73 .
the noise is gaussian distribution with @xmath74 and the snr of the system is defined as @xmath75 .
simulation results are measured in terms of bit error rate ( ber ) of @xmath13 at the relay node since the broadcast phase is straightforward . in figure 2 ,
we plot the ber performance of different schemes .
as shown in the figure , the proposed mimo pnc schemes always outperform their counterparts .
specifically , our vblast pnc outperforms vblast nc by about 0.5 db at a ber of @xmath76 ; the sorted vblast pnc outperforms sorted vblast nc by about 1db . in sorted vblast
, the more performance improvement mainly comes from the fact that the average value of @xmath62 is smaller and the interference to be cancelled , @xmath77 , is smaller .
in this paper , a novel signal detection and network encoding scheme , vblast pnc , is proposed to extract @xmath13 at the relay node in mimo twrc .
the basic idea is that the relay node first uses partial interference cancellation to obtain @xmath53__kx__@xmath78during vblast detection process and then converts it to @xmath13 with pnc mapping . with partial interference cancellation
, error propagation effect is mitigated and the performance is significantly improved .
the simulation results verified the performance advantages of vblast pnc under the setting of random rayleigh fading channel coefficients .
our scheme is of great interest in practice since only csir , symbol lever synchronization and low complexity are needed .
this work was partially supported by nsfc ( no . 60902016 ) , nsf guangdong ( no . 10151806001000003 ) , and nsf shenzhen ( no . jc201005250034a ) .
y. rong , x. tang and y. hua , `` a unified framework for optimizing linear nonregenerative multicarrier mimo relay communication systems '' , ieee transactions on signal processing , pp : 4837 - 4851 , dec .
2009 s. zhang , s. c. liew , and p. p. lam , `` physical layer network coding , '' in proc .
acm mobicom06 : the 12th annual international conference on mobile computing and networking , pages 358365 , new york , ny , usa , 2006 . | for mimo two - way relay channel , this paper proposes a novel scheme , vblast - pnc , to transform the two superimposed packets received by the relay to their network coding form .
different from traditional schemes , which tries to detect each packet before network coding them , vblast - pnc detects the summation of the two packets before network coding .
in particular , after firstly detecting the second layer signal in 2-by-2 mimo system with vblast , we only cancel part of the detected signal , rather than canceling all the components , from the first layer
. then we directly map the obtained signal , summation of the first layer and the second layer , to their network coding form . with such partial interference cancellation
, the error propagation effect is mitigated and the performance is thus improved as shown in simulations .
shell : bare demo of ieeetran.cls for journals multiple input multiple output , physical layer network coding , two way relay channel , vblast . |
psr b1706@xmath044 , discovered by @xcite is among the most interesting pulsars for study at high energies .
it is one of a handful of pulsars detected by egret in gev @xmath1-rays .
it is quite similar to the vela pulsar with a characteristic age @xmath2yr and a spindown luminosity of @xmath3 erg / s , but is @xmath4 more distant at @xmath5kpc .
early _ chandra _ hrc / acis data provided a first detection of x - ray pulsations and showed a compact @xmath6 surrounding pulsar wind nebula ( pwn ) @xcite .
more recent _ xmm - newton _
spectroscopy @xcite has provided improved measurements of the x - ray spectrum and pulsations .
early claims that the pwn is detected in tev @xmath1-rays @xcite have not been supported by recent hess observations @xcite .
psr b1706@xmath044 is superposed on a radio - bright spur of the supernova remnant [email protected] , which has a similar , albeit unreliable , @xmath7 distance of @xmath8kpc @xcite .
the pulsar dm gives a distance of @xmath9kpc in the @xcite model .
@xcite have argued for an association . in particular
, they found a faint southern extension of the snr , which would place the pulsar within the full snr boundary .
they also noted an approximately n - s elongation of the x - ray pwn , pointing roughly back to the snr center and argued that this would represent a trailed nebula .
the required velocity for travel from the approximate geometric center of the snr , about @xmath10 away , was @xmath11km / s where @xmath12 is the pulsar age in units of @xmath13yr .
there are , however , some challenges to this snr association .
@xcite in an hi absorption study of the pulsar found velocity components setting lower and upper bounds for the distance of @xmath14kpc and @xmath15kpc .
however a prominent hi emission feature seen in the bright limb of [email protected] at @xmath16km / s is not seen by koribalski et al . in the absorption spectrum of the pulsar , suggesting that it lies in front of the snr .
also scintillation studies @xcite suggest a low transverse velocity for the pulsar , @xmath17km / s . this estimate has been supported by more recent scintillation measurements ( johnston , priv .
thus the distances of the pulsar and snr are still fairly uncertain .
we adopt here a generic distance of 3kpc in the discussion that follows , but carry through the scaling to show the distance dependence .
@xcite re - examined the ao1 50ks hrc and @xmath18ks acis - s exposures , and found that the compact pwn could be well modeled as an equatorial torus + polar jets , rather similar to the structures seen around the young crab and vela pulsars .
they found that the brightest arc of pwn emission lay _ behind _ the pulsar s inferred motion from the snr center , while a polar jet extended well in front of the pulsar position .
this is difficult to reconcile with a bow shock / trail interpretation . on the other hand ,
the pwn symmetry axis did indeed point back to the snr center , suggesting that a more careful evaluation of the connection was in order .
this is particularly interesting , since comparison of the pwn symmetry (= pulsar spin ) and proper motion axes can constrain the origin of pulsar birth kicks @xcite .
we have obtained a deeper 100ks acis - i exposure of psr b1706@xmath044 and surroundings .
the x - ray exposure coverage is compared to the overall geometry of [email protected] in figure 1 .
together with new atca radio continuum imaging we are able to study the rich structure in this pwn and further constrain its connections with the snr .
psr b1706@xmath044 was observed with the _ chandra _ acis - i array ( 4 acis - i chips along with the s3 and s4 chips ) on february 1 - 2 , 2004 with standard imaging ( 3.2s te ) exposures .
the ccd array was operated in ` very faint ' ( vf ) mode , allowing improved rejection of particle backgrounds .
the total live - time was 98.8ks and no episodes of strong background flaring were observed .
hence all data are included in our analysis .
the pulsar was positioned near the standard aimpoint of the i3 chip and all observing conditions were normal .
we have also compared our new exposure with the archival ( february 3 , 2001 ) 14.3ks acis - s3 exposure ( obsid 0757 ) . as usual the backside illuminated s3 chip suffered more from particle background and after cutting out periods of background flares , 11ks of clean exposure remained .
all analysis was performed using ciao 3.2 and caldb 3.0.0 , including automatic correction for the acis qe degradation .
these data were nearly free of pile - up ; the maximum pixel counts at the pulsar position indicate only @xmath19% pile - up while the best fit model for the point source has an expected pile up fraction of @xmath20% . for sources with low pile - up we can maximize the spatial resolution of the acis image by removing the standard pixel randomization and applying an algorithm correcting the position of split pixel events @xcite .
this decreases the on - axis psf width in our data set by @xmath21% .
these data are compared with radio observations of the pwn .
data for the radio maps shown here were collected at the australia telescope compact array ( atca ) in narrabri ( latitude -30.3@xmath22 ) @xcite . for the 1.4ghz map in figures 1 and 2a , the data acquisition and analysis
are described in @xcite . for the image contours in figure 3a ,
the data first presented in @xcite were re - imaged including the 6 km baselines and uniform weighting to highlight the high resolution features .
the restoring beam size is @xmath23 . to show the nebular structure ,
an 11mjy point source psf has been subtracted at the position of the pulsar .
two maxima appear flanking the pulsar position .
these are unlikely to be artifacts due to pulsar variability , as diffractive scintillation for this pulsar is particularly weak @xcite .
since the data were collected in five sessions , spread over more than a year , it is in principal possible for slow refractive scintillation to change the pulsar flux between epochs and distort its psf .
however , each epoch used @xmath2412h of integration , so any residual epoch psf should be close to circularly symmetric , in contrast to the structure near the pulsar which is clearly bipolar . further observations , with pulsar binning , have been requested to confirm this result . for the 4.8ghz map in figure 2b , observations were made at 4.8- and 8.6ghz with the array in the standard configurations 0.75a , 1.5a and 6a on 06 jan , 16 feb and 11 apr 2002 .
the maximum and minimum baselines for the 4.8-ghz data were 1 and 100 k@xmath25 ( angular resolutions of @xmath26 to @xmath27 ) for a total of 26 hours observation .
in all cases we observed the two frequencies with bandwidths of 128 mhz .
we used the atnf correlator mode that divides each integration s data into separate phase bins spanning the pulsar period .
this firstly allowed the strongly pulsed point source flux to be excluded from the image and secondly allowed us to self - calibrate using the relatively strong point source flux from the pulsar .
after data editing and calibrating we inverted the image with a _ uv_-taper of @xmath28 and deconvolved it with the full polarization maximum entropy task pmosmem in miriad .
the most important test of the snr association would , of course , be a direct astrometric proper motion . with a 1.4ghz flux of @xmath2411mjy ,
psr b1706@xmath044 is relatively bright . as such it is suitable for phase referenced vlbi astrometry , if an in - beam reference could be found .
unfortunately searches for phase references adequate for australian long baseline array ( lba ) and us vlba experiments have not detected comparison sources with compact fluxes greater than @xmath29mjy .
attempts were made at external phase reference vlba astrometry .
however at 1.4ghz , the nearest known reference source ( 2.5@xmath22 away ) was scatter broadened to @xmath30mas .
with the strong ionosphere at such low elevation , the next nearest known source ( 10@xmath22 away ) is too distant for effective calibration .
since the pulsar spectrum is steep , an attempt at vlba astrometry at 5ghz was also unsuccessful ; at this low elevation the system temperature was @xmath31 nominal and only six vlba antennae could be used , reducing the sensitivity to @xmath32 of nominal .
so unfortunately we have only tied - array astrometry at present .
even if the pulsar does travel from the geometric center of [email protected] , the expected proper motion is only @xmath33mas / yr ; the existing time base of vla / atca imaging does not yet allow a serious constraint on this motion .
we must conclude that a direct proper motion measurement awaits substantially increased ( ska or evla ) capabilities and a long - duration , large base - line experiment . to show the diffuse emission surrounding psr b1706@xmath044 we plot ( fig .
2a ) a 1 - 7kev image with point sources removed ( except the pulsar ) .
these data are exposure corrected to minimize the chip gaps and heavily smoothed on a @xmath34 scale .
the diffuse emission is an edge - brightened , radius @xmath35 cavity surrounding the pulsar with a faint extension to the west .
contours of the 1.38ghz radio map show good correlation with the radio emission in the bar crossing [email protected] ( fig .
1 ) . we will refer to this structure as the ` nebula ' .
gaussian kernel smoothing .
contours are from the 1.38ghz radio map of figure 1 .
right : @xmath36 gaussian smoothed image of the pwn with an overlay of the core of the radio nebula from a 4.8ghz atca image ( contours at 0.4 , 0.8 , 1.0 , 1.1, ... 1.6 mjy / beam ) ; the resolution is @xmath34 and the image rms is 0.2mjy / beam .
, title="fig : " ] gaussian kernel smoothing .
contours are from the 1.38ghz radio map of figure 1 .
right : @xmath36 gaussian smoothed image of the pwn with an overlay of the core of the radio nebula from a 4.8ghz atca image ( contours at 0.4 , 0.8 , 1.0 , 1.1, ... 1.6 mjy / beam ) ; the resolution is @xmath34 and the image rms is 0.2mjy / beam .
, title="fig : " ] moving in toward smaller scale , in 2b we show a 1 - 7kev image , smoothed with a @xmath36 gaussian .
point sources have not been removed .
this shows that the cross structure fit by @xcite extends across @xmath37 .
narrow x - ray jets , which we refer to here as the ` outer jet ' ( extending south ) and ` outer counter jet ' ( extending north ) start @xmath6 from the pulsar and continue to @xmath38 .
bracketing these is faint diffuse x - ray emission which we will call the ` equatorial pwn ' . for comparison
we draw contours of a 4.8ghz atca image with @xmath39 restoring beam .
these observations have the pulsar ` gated out ' and show that the radio pwn has a hollow center bracketing the equatorial pwn .
diffuse radio peaks are , in fact , seen just east and west of this x - ray structure .
finally , we show in figure 3 a lightly smoothed image of the central region of the pwn , stretched to bring out the faint outer jets .
the contours are drawn from a 1.38ghz atca image , where the 6-km baselines have been weighted to produce a @xmath23 restoring beam .
a point source psf has been subtracted at the pulsar position .
two local radio maxima with peak fluxes @xmath40mjy and @xmath41mjy bracket the ` torus ' structure .
the radio then shows a sub - luminous zone surrounding the ` equatorial pwn ' ; beyond @xmath38 the radio brightens again , as in figure 2b .
no emission appears along the ` outer jets ' .
indeed there appear to be evacuated channels in the radio emission , but improved s / n and resolution are needed to probe this sub - mjy structure .
the frames on the right show the innermost region of the pwn with the best - fit torus + inner jet model ( 2.3 ) .
gaussian smoothing ) of the pwn , stretched to show the outer jets and the equatorial pwn .
the contours ( 0.3 , 0.4 , 0.5 , 0.6 , 0.9 , 1.2 , 1.5mjy / beam ) are from the 1.38ghz atca image , which has a map rms of 0.2 mjy / beam and a resolution of @xmath23 .
the pulsar point source has been subtracted .
two peaks of radio emission bracket the torus .
these lie in a cavity which surrounds the equatorial pwn and jets .
middle : grey scale image of the inner pwn ( above ) and of the best - fit point source+torus+(inner)jets model ( table 1 ) to the same scale(below ) . at the far right ,
the residual image is shown , with contours indicating the excess counts above ( north of ) the torus .
, title="fig : " ] gaussian smoothing ) of the pwn , stretched to show the outer jets and the equatorial pwn .
the contours ( 0.3 , 0.4 , 0.5 , 0.6 , 0.9 , 1.2 , 1.5mjy / beam ) are from the 1.38ghz atca image , which has a map rms of 0.2 mjy / beam and a resolution of @xmath23 .
the pulsar point source has been subtracted .
two peaks of radio emission bracket the torus .
these lie in a cavity which surrounds the equatorial pwn and jets .
middle : grey scale image of the inner pwn ( above ) and of the best - fit point source+torus+(inner)jets model ( table 1 ) to the same scale(below ) . at the far right ,
the residual image is shown , with contours indicating the excess counts above ( north of ) the torus .
, title="fig : " ] gaussian smoothing ) of the pwn , stretched to show the outer jets and the equatorial pwn .
the contours ( 0.3 , 0.4 , 0.5 , 0.6 , 0.9 , 1.2 , 1.5mjy / beam ) are from the 1.38ghz atca image , which has a map rms of 0.2 mjy / beam and a resolution of @xmath23 .
the pulsar point source has been subtracted .
two peaks of radio emission bracket the torus .
these lie in a cavity which surrounds the equatorial pwn and jets .
middle : grey scale image of the inner pwn ( above ) and of the best - fit point source+torus+(inner)jets model ( table 1 ) to the same scale(below ) . at the far right ,
the residual image is shown , with contours indicating the excess counts above ( north of ) the torus .
, title="fig : " ] the overall geometry of the pwn is strongly reminiscent of that surrounding the vela pulsar .
in particular @xcite have described a series of acis images of the vela nebula which show a torus - like structure , an inner jet and counter jet and a faint narrow outer jet system .
this imaging sequence showed that the vela outer jet , which is patchy and strongly bent , varies dramatically on timescales of days to weeks .
apparent motion of blobs within the jets suggests mildly relativistic bulk velocities and strong instabilities . for psr
b1706@xmath044 our single sensitive image does not let us comment on variability .
however we argue that the relatively straight and narrow jets , @xmath42 longer than those of vela , and symmetric pwn structure are a consequence of a static uniform external environment and a low pulsar velocity . at 1.4 - 8.5ghz @xcite
have found that the vela pwn has two bright patches bracketing the x - ray torus and jets , in a structure quite similar to that in figure 2b .
polarization imaging of the vela radio structure suggests that these two patches represent the limbs of a toroidal b field structure .
this implies that the rotation axis controls the pwn symmetry to large radii .
following @xcite we have fitted our new acis image to a point source psf , doppler boosted equatorial torus , polar jets and uniform background .
the fitting minimizes residuals using a poisson - based likelihood function .
monte carlo simulations of poisson realizations of the best - fit model are in turn re - fitted to generate statistical errors and their co - variance matrices .
table 1 contains the best - fit values .
the torus radius and axis inclination and position angles are @xmath43 , @xmath44 and @xmath45 , respectively .
see @xcite for the definition of the other parameters and the details of the fitting technique . in table 1 the inner jet / counter - jet are constrained to lie along the torus axis in the fits .
= .torus fit parameters with 1@xmath46 statistical errors [ cols="^,^,^,^,^,^,^,^,^,^,^",options="header " , ] @xmath47 0.5 - 8kev fluxes in units of @xmath48 @xmath49 held fixed @xmath50 simultaneous @xmath51 fit .
[ pwnspec ] note that there is significant softening of the extended emission as one progresses to larger scales and that the jet components appear to be the hardest of all .
this is certainly consistent with the idea that the central pulsar supplies fresh energetic electrons and that synchrotron burn - off increasingly softens the spectrum as older populations are viewed in the outer pwn .
again this trend is common in the well - measured young pwne .
allowing the photon index to vary for the different nebula components , the best fit to a global absorption value for the extended emission gives us our fiducial @xmath52 .
this is consistent with free - fit values for the point source , but given complexities of the composite thermal+power law model , we consider the nebular fit value more robust .
note that with dm=75.7 @xmath53 , the @xmath54 for this sight - line is large , but not unprecedented for low @xmath55 pulsars .
this is also consistent with the @xmath56 absorption measurements and a fiducial snr distance @xmath8kpc , given the appreciable uncertainties .
a number of authors have discussed the evolution of a pwn within an expanding supernova remnant .
for example , @xcite and @xcite describe the early evolution when the supernova ejecta are in free expansion .
later , after the remnant interior is heated by the passage of the reverse shock , the pwn evolves within the sedov phase supernova remnant whose radius is @xmath57 for an explosion energy @xmath58 in a @xmath59 medium of density @xmath60 .
psr b1706@xmath044 has a characteristic age @xmath61yr with @xmath62 , so [email protected] should be safely in the sedov phase with an expected angular size @xmath63 for a supernova releasing energy @xmath64 in an external medium density @xmath65 , at a true age @xmath66y at a distance @xmath67kpc .
the observed size then implies @xmath68 , requiring a fairly energetic explosion for @xmath69kpc . during the sedov phase the interior pressure is @xmath70 and is relatively constant away from the snr limb .
the pulsar blows a wind bubble within this snr interior , whose radius is @xmath71 for a pwn bubble energy @xmath72 @xcite .
although the accuracy of this dependence of pwn radius on pulsar injection energy has been questioned @xcite , we adopt it for the following estimates . with the observed ratio of radii , @xmath73
, we obtain @xmath74 , i.e. this pwn has quite low internal energy .
this is also reflected in the low radio and x - ray fluxes .
together we use these estimates , the observed size of the snr , equation ( 1 ) and the measured @xmath75 and @xmath76y to write @xmath77 . now , if the pwn is adiabatic and we assume spindown with constant @xmath78 and braking index @xmath79 from an initial period @xmath80 , we find that the total energy in the plerion is @xmath81 { \dot e } \tau_c$ ] . then , setting @xmath82 and eliminating the true age @xmath83 using @xmath84 $ ] for magnetic dipole spindown , we obtain a constraint on the initial spin period @xmath85 ^ 3/(p_0/p)^2 = 0.68\ , n_0 d_3 ^ 5 \eqno(3)\ ] ] which has a solution of @xmath86ms for d=3kpc , and @xmath87ms for d=2kpc .
the corresponding true ages are 0.61@xmath88 ( @xmath89y ) and 0.38@xmath88 ( @xmath90y ) , respectively .
these numerical values are for @xmath91 and the density dependence from equation ( 3 ) is quite weak .
@xcite present a similar sum for @xmath80 , assuming a known @xmath58 ; the above formulation emphasizes the sensitivity to the poorly known @xmath92 .
note that with the large implied initial period , the integrated pwn energy is quite comparable to the present spin energy , with @xmath93 at d=3kpc and @xmath94 at d=2kpc .
so the spindown luminosity is roughly constant in the adiabatic phase and the pwn growth is closer to @xmath95 than to the @xmath96 law appropriate for impulsive energy injection @xcite . inside this wind bubble , the sedov interior pressure confines the pwn , giving rise to a termination shock at @xmath97 which results in @xmath98 .
if we apply the snr estimate for @xmath58 above , this becomes @xmath99 .
then , using @xmath100 and applying the age estimate following equation ( 3 ) we get @xmath101 ( d=3kpc ) or @xmath102 ( d=2kpc ) .
these estimates are reasonably consistent with the observed @xmath103 torus radius , especially since an equatorially concentrated flow should have a stand - off distance 1.5 - 2@xmath104 this spherical scale .
the polar jets can have an initial shock at somewhat larger angle , with the resulting pitch angle scattering illuminating the jets somewhat further from the pulsar .
of course , this bubble is offset from the center of [email protected] at @xmath105 ( figure 1 ) .
this is inside the @xmath106 where @xcite note that the increasing density causes the pulsar to be supersonic , so a bow shock should not have yet formed .
these authors however compute numerical models of a fast moving pulsar in a snr interior . as the pulsar moves
, the pwn should become highly asymmetric with a ` relic pwn ' at the snr center and the pulsar placed near the leading edge of the pwn ; see @xcite figures 7 and 8 .
we see no pwn structure near the geometric center of [email protected] and , if the ` bubble nebula ' is identified with the shocked pulsar wind , the pulsar is certainly not offset from its center along the proper motion axis ( away from the snr center ) .
so these models are an inadequate description of [email protected] . from figure 2 ,
the pulsar is well centered in the bubble nebula , with any offset from its center along the axis to the snr substantially less than @xmath107arcsec .
thus @xmath108 and the pulsar can not have moved far from the explosion center .
this is , of course , consistent with the scintillation results .
we can reconcile the symmetric pwn with the offset snr shell if we assume that the pulsar progenitor exploded toward the edge of a quasi - spherical cavity .
one scenario ( also posited by gvaramadze 2002 , bock & gvaramadze 2002 ) that can associate the low velocity pulsar with [email protected] is to assume that the progenitor star had a stellar wind of mass loss rate @xmath109/yr and wind speed @xmath110cm / s over @xmath111y , typical of the @xmath112 stars that dominate the pulsar progenitors @xcite .
this evacuates a stellar wind bubble of size @xmath113 during the main sequence lifetime , the star moving at @xmath114km / s travels @xmath115 and so it can easily traverse its wind bubble .
thus , one can imagine an off - center supernova in a nearly symmetric stellar wind bubble of radius @xmath116 : the supernova blast wave expands to fill the bubble , passing to the sedov phase near its present radius .
the supernova produces a neutron star with little or no kick , placing the pulsar near its present position .
this has the added advantage of accommodating the rather large snr size with a more modest energy of a few@xmath117erg .
the pwn energy and size estimates above would then be somewhat amended ; this would require a careful numerical simulation .
for the reasons detailed in the introduction , it is not yet clear that psr b1706@xmath044 and [email protected] are associated .
so for completeness we can consider the case when the shocked pulsar wind blows an adiabatic bubble in a static , low @xmath118 external medium @xcite .
if we assume that the pulsar was born ( sans snr ) or entered a confining region of the ism @xmath119y ago and that since then it has been spinning down at the present energy loss rate , we find that it will blow a bubble of angular size @xmath120 these estimates change somewhat for a pulsar born at @xmath121 ; since we are not making the association with the snr [email protected] , we can make no estimate of the initial spin period . as first noted by @xcite , the @xmath122 wide radio spur across the face of [email protected]
has the approximate scale of such a ` bubble nebula ' .
if the pwn stays unmixed ( relativistic ) then the interior of the bubble will have a pressure @xmath123 . in turn , the torus termination shock in this medium is at @xmath124 if ( e.g. through rayleigh - taylor instabilities ) the pulsar wind is well mixed with the swept up gas , the adiabatic thermal pressure would be @xmath125 larger .
interestingly , the angular scales for this scenario are also reasonably compatible with the observed torus and bubble nebula size .
of course this scenario leaves open the question of the pulsar origin .
again the pulsar would need to have a quite low velocity to produced the observed symmetry . turning to the spectral results , we note that @xcite fit a correlation between spindown energy and the psr+pwn luminosity : @xmath126 .
for the psr b1706@xmath044 parameters this predicts a flux @xmath127 .
the observed 2 - 10kev flux is in fact @xmath128 , even including the outer ` bubble nebula ' ; without this component it is half as large .
these correlations are not very accurate , but this does imply that the psr b1706@xmath044 pwn is substantially under - luminous for any distance less than 3kpc .
@xcite has derived correlations between the pulsar spindown power and the pulsar / pwn spectral indices .
his relation predicts @xmath129 ( substantially smaller than our power law index @xmath130 ) and @xmath131 ( not inconsistent with the values measured for the torus and equatorial pwn ) .
as described in 2.4 , the spectrum softens appreciably from the central torus to the outer bubble nebula .
this suggests increased aging of the synchrotron population .
figures 2 and 3 show that the bulk of the radio emission lies in the ` bubble nebula ' region .
so we can take the radio flux and spectral index from @xcite and compare with our nebula x - ray flux ( fig .
4 ) . comparing the radio spectral index
@xmath132 with the best fit x - ray index @xmath133 , shows a break quite close to the @xmath134 expected from synchrotron cooling .
the extrapolated intersection of these power laws gives a break frequency of log[@xmath135=12.2^{+0.9}_{-1.1}$ ] .
for the fiducial pulsar age of @xmath136y , this corresponds to a nebula field of @xmath137 g .
note that the magnetic pressure from this ( photon flux - weighted ) average field is @xmath138 , somewhat larger than the nebula pressure estimated from its radius .
this may indicate field compression in the nebula limb .
in general , if the mean nebula field is @xmath139 g for a nebula of angular radius @xmath140 , the total nebula field energy is @xmath141erg .
this is comfortably less than the present spin energy @xmath142erg , so the nebula can be easily powered even if the pulsar was born close to its present spin period .
we find that this cooling break field is substantially larger than the equipartition field of @xmath143 g inferred for the radio and x - ray emitting populations ( also the minimum equipartition nebula energy @xmath144erg is substantially smaller ) .
the cooling break field can also be compared to that expected from simple radial evolution of the pulsar surface field : if this field @xmath145 g falls off as @xmath146 to the light cylinder , then as @xmath147 to the wind shock where it is compressed we get @xmath148 mg . if it continues to fall off as @xmath147 beyond this we get a field at the limb of the bubble nebula of @xmath149 g .
so the best we can do is to infer a mean nebular field @xmath150 the equipartition value , with some generation of new field beyond the torus wind shock .
the energetic requirements for this field , required to match the @xmath151 cooling break , are comfortably less than the energy available from psr b1706@xmath044 .
these field estimates are consistent with the non - detection of tev ics flux from this source @xcite .
the narrow outer jets also have a power - law spectrum and are almost certainly synchrotron - emitting . for a reasonable @xmath153 mg field ,
the observed x - rays of @xmath154kev require substantial @xmath155 energies , with @xmath156 near the radiation - reaction limited primary lorentz factor inferred for many polar cap @xcite and outer magnetosphere @xcite pulsar models .
since the jet is narrow , confinement of these pairs imposes a ( not very restrictive ) lower bound on the jet field @xmath157 where the maximum observed jet photon has @xmath158kev and the observed outer jet half - width is @xmath159arcsec .
a more restrictive upper limit on the mean jet field comes from the observation that the jets do not soften noticeably before their end @xmath160arcsec from the pulsar .
if we assume a jet bulk speed @xmath161c , then arguing that the flow time is shorter than the synchrotron cooling time gives us the limit @xmath162 when the observed jet spectrum has an energy index of @xmath163 , we infer a power - law spectrum of @xmath155 in the jet @xmath164 , with @xmath165 .
we can then make an estimate of the minimum jet luminosity , i.e. at ` equipartition ' when @xmath166 .
given the observed combined outer jet luminosity ( 0.5 - 8kev ) @xmath167 and emitting volume @xmath168 with the angles in arcsec , we can estimate the equipartition field for an isotropic plasma as @xmath169^{2/7 } \eqno ( 9)\ ] ] where the observed photon spectrum runs from @xmath170 to @xmath171 .
for the observed flux this gives @xmath172^{-2/7}\ , { \rm g } , \eqno ( 10)\ ] ] where @xmath173 is a weak function of @xmath174 .
the corresponding minimum energy flux for the outer jet is @xmath175 where the jet bulk velocity is @xmath161c .
this is @xmath176 per jet and will , of course , be larger if the jet flow includes ions . interestingly ,
if the pulsar couples roughly isotropically to the pwn , then the corresponding fraction of the outflow should subtend a half angle of @xmath177 .
this is somewhat smaller than the angle subtended by the inner jets , but @xmath42 larger than the @xmath178 width of the ends of the jet
there is substantial collimation of the jet energy flux .
we have argued that a low psr velocity can explain the symmetry of the pwn .
the central location of the pulsar and spherical post - shock flow may also allow the equatorial toroidal structure and polar jets to propagate undisturbed to large radii .
we do , however , measure a small misalignment of the outer jets , corresponding to a deflection of @xmath179 for each .
if we imagine a pressure acting along the jet s @xmath38 length , then the required perturbation is @xmath180 , where @xmath181 is the jet s cross sectional area .
this is only @xmath182 of the total pressure in the nebula
. it could be due to ram pressure if the shocked nebular medium flows to the west at @xmath183km / s .
our x - ray measurements have established the pwn symmetry axis , presumably reflecting the pulsar spin axis , to very high precision .
unfortunately our original goal of relating this to the proper motion axis remains unfulfilled .
it is true that the torus symmetry axis points roughly toward the center of [email protected] , confirming the estimates from earlier _ chandra _ data .
however , the pwn symmetry about the pulsar and the low scintillation velocity suggest a very low transverse speed @xmath184km / s , which would preclude a birth site as distant as the snr center .
this low speed makes a direct proper motion challenging , but allows latitudinal asymmetries in the pwn flow to propagate undisturbed to fairly large radius , where they can be imaged with _
thus , study of this pwn offers some good opportunities to probe outflow dynamics and jet collimation .
study of this , and similar , pwne may prove useful electrodynamic analogs of the @xmath185 more powerful agn jets . a viable scenario for maintaining the [email protected]/psr b1706@xmath044 association posits a supernova event near the present pulsar site , with the remnant inflating a pre - existing off - center cavity .
however , the residual ( small ) proper motion could then have any direction .
in fact , the faint extension of the pwn ( bubble nebula ) to the west and the increased radio surface brightness to the east might suggest a rather slow pulsar motion at pa @xmath186
. this would be nearly orthogonal to the torus symmetry axis . with the large @xmath80 estimated here , this
could be construed as suggesting poor rotational averaging of a birth kick @xcite .
so the pwn / snr geometry offers both aligned and orthogonal axes .
only a sensitive astrometric campaign can detect or limit the pulsar motion and resolve this ambiguity .
this work was supported in part by nasa grants sao g04 - 5060x and nag5 - 13344 .
we thank the referee for several careful readings , which resulted in substantial changes .
we also thank r.n .
manchester who provided a current pulsar ephemeris for the vlbi experiment . | we report on _ chandra _ acis imaging of the pulsar wind nebula ( pwn ) of the young vela - like psr b1706@xmath044 , which shows the now common pattern of an equatorial wind and polar jets .
the structure is particularly rich , showing a relativistically boosted termination shock , jets with strong confinement , a surrounding radio / x - ray pwn and evidence for a quasi - static ` bubble nebula ' .
the structures trace the pulsar spin geometry and illuminate its possible relation to snr [email protected] .
we also obtain improved estimates of the pulsar flux and nebular spectrum , constraining the system age and energetics . |
over the past five years , direct imaging surveys for extrasolar planets have discovered a small but significant number of ultra - low mass companions ( henceforth ulmcs , masses @xmath320@xmath4 ) at @xmath550 au separations from their primaries .
these objects have estimated masses that are in the same range as radial velocity or transiting planets with @xmath65au separations , which have a continuous mass distribution up to @xmath020@xmath4 , then a gap in mass until arguably star - like objects are found at @xmath760m@xmath8 @xcite .
for this reason , ulmcs are often called `` planetary mass '' companions .
the prototypical wide ulmc , 2m1207 - 3933 , consists of a 48 @xmath9 companion located @xmath050 au away from a 10 myr old brown dwarf @xcite . since its discovery ,
half a dozen other ulmcs have also been reported , most of which orbit much higher - mass primaries ( @xmath00.52.0 @xmath10;e.g .
@xcite ; @xcite ; @xcite ) .
some of these systems appear to be genuinely scaled - up versions of our own solar system , with planets that are consistent with formation in a disk .
one example is the companion to fomalhaut , which is coplanar with its debris disk .
another is the hr 8799 , which has multiple planetary companions of similar mass .
other cases like chxr 73 and 1rxs j160929.1 - 210524 are more ambiguous since their orbital radii are even wider , and it is unclear whether they lie in the original plane of planet formation .
such companions could very well form like planets within a circumstellar disk or like binaries from the collapse of a molecular cloud .
these ulmcs pose a significant challenge to existing models of planet and binary formation .
their orbital radii are so large that the core accretion timescale ( @xmath7100 myr at 100 au ; pollack et al .
1996 ) should be much longer than the typical protoplanetary disk dissipation timescale ( @xmath03 - 5 myr ; @xcite ; @xcite ; @xcite ) .
some of the closer companions could potentially form on the gravitational instability timescale ( e.g. * ? ? ?
* ) , which can be very short at intermediate radii ( 10100au ) , but gravitational instability in disks has not been extensively modeled at @xmath5100 au , much less at @xmath0300 au .
these companions could represent the extreme end of the binary mass function , which appears to be linearly - flat ( with all companion masses being equally probable , @xcite ) well into the substellar regime .
however , it is unclear whether this trend could extend to planetary masses ( by which we mean @xmath620m@xmath1 in this paper ) .
these companions have masses near the opacity - limited minimum mass @xcite , and unless their formation occurred exactly as the circumstellar envelope was exhausted , then they should have quickly accreted enough mass to become high - mass brown dwarfs or stars . in this paper , we report the discovery of two ulmcs in the upper scorpius ob association , one of which was independently discovered by @xcite .
the survey consists of a subset of the aperture - masking interferometry sample reported by @xcite .
we describe our observations and data analysis techniques in section [ sectobs ] , and in section [ sectfollowup ] , we report the detections and detection limits from our survey . in section [ sectfreq ] , we report the first measurement of a frequency for ulmcs around young stars
. finally , in section [ sectdiscuss ] , we discuss the implications for possible formation mechanisms for ulmcs .
nearby ( @xmath3200pc ) young ( @xmath320myr ) stars have been the object of numerous high - resolution imaging campaigns over the past several decades .
these observations have included lunar occultation ( e.g. @xcite ) , speckle interferometry ( e.g. @xcite ) , adaptive optics ( ao ) imaging ( e.g. @xcite ) , and most recently non - redundant masking ( nrm ) inteferometry @xcite .
the earlier techniques of lunar occultaion and speckle were only sensitive to the presence of bright , stellar - mass , companions , and only recently did ao imaging and nrm interferometry managed to probe within the brown dwarf regime , going as far as sampling the top of the planetary mass regime . in k08 , we reported the results of one such survey of young stars in the upper sco ob association that used a combination of conventional ao imaging and nrm - interferometry .
in addition to the results for stellar and brown dwarf companions , that paper also reported detection limits for ulmcs at small separations ( within 50 au ) where the probability of background star contamination was negligible . in this work ,
we report the corresponding analysis for candidate companions at wide separations , including multi - epoch follow - up imaging for three candidates , of which two appear to be associated and of approximately planetary mass .
table 4 of k08 lists the ao imaging observations conducted with the pharo camera at the palomar 200 " telescope and the nirc2 camera at the keck - ii 10 m telescope .
we found that 10 out of these 62 targets had stellar binary companions at separations of @xmath00.255.0 , which should mean that the majority of additional faint companions are not dynamically stable in this range of separation ; these were omitted from our sample .
in addition , we omitted the few targets with spectral types of later than m2 ( @xmath11 ) so as to work with a single mass range of approximately `` solar - type '' stars , leaving 49 targets in consideration .
all observations used the smallest pixel scales ( 10 mas / pix with nirc2 and 25 mas / pix with pharo ) in order to achieve the best psf sampling .
we used a @xmath12 filter at palomar and the @xmath13 filter at keck , which yielded diffraction - limited resolutions of 100mas and 50mas , respectively .
the keck observations used the narrowband filter despite the penalty in sensitivity , because the brighter primary stars in our sample would have saturated the detector within the minimum exposure time .
much of this sensitivity was regained by using more fowler samples per frame , which significantly reduces the high read noise of nirc2 s detector .
these observations used relatively short total integration time , of the order of a minute , in comparison with other ao imaging surveys ( e.g. @xcite ) , resulting in limiting magnitudes of @xmath141517 for the companions . given the young age of these upper sco targets ( @xmath05 myr ) , even these shallow observations were able to reach the planetary mass regime ( at 2 arcsec separations , @xmath314 @xmath9 for 48 targets and @xmath37 @xmath9 for 42 targets ) . the detections and detection limits were derived using methods we previously described in @xcite and kraus & hillenbrand ( 2010 , in prep ) . since the detection limits at more than
a few @xmath15 are driven by speckle noise , detections and detection limits were determined from the coadded image stacks by placing a large number of photometric apertures ( with diameter @xmath16@xmath17 ) around each target , then measuring the mean and standard deviation of the brightness distribution for all apertures in concentric rings around the primary . for each ring
, we identified all candidate detections with significance @xmath55@xmath18 above the mean brightness , then compared those detections to other stars taken on the same night to identify and reject the quasistatic speckles and diffraction spikes that are seen in common for many targets .
all candidate detections that could not be identified as psf artifacts were then adopted as genuine sources and hence as candidate bound companions , while the 5@xmath18 limits were adopted as our formal detection limits . at large separations ,
the detection limit from this algorithm converges to the sky background limit , and at these separation ( @xmath72 arcsec ) , we used a 10@xmath18 limit , because of the large number of pixels in this regime .
we found no candidate companions near the detection limits in the speckle - limited regime , but there are many candidate companions in the sky - limited regime .
we determined the photometry and astrometry for these sources using the methods described in k08 and @xcite .
to briefly summarize , we measured astrometry and aperture photometry for each source with respect to the known usco member using the iraf task daophot @xcite ; all measurements were conducted using apertures of 0.5 , 1.0 , and 2.0 @xmath15 , and then the optimal aperture was chosen to maximize the significance of the detection ( given the competing uncertainties from the sky background and the poisson noise for the source itself ) . in order to estimate the uncertainties from the data , we analyzed the measurements in individual frames and then combined those measurements to estimate the mean and standard deviation .
we then accounted for the systematic uncertainty in the plate scales and distortion solutions of pharo ( 0.3% ; section 2.3 ) and nirc2 ( 0.05% ; ghez et al .
2008 , @xcite ) by adding those terms in quadrature with the observed scatter .
most of the new candidate companions were identified with pharo ; as we describe below , its astrometric calibration is not yet well - understood and might be further improved with more calibration observations , but our systematic uncertainties should account for this effect .
many of the wider ( @xmath55 ) companions are also affected by anisoplanatism , yielding systematically low brightness estimates for aperture photometry .
a correction of the photometry would require a detailed knowledge of the atmospheric turbulence profile and isoplanatic patch size , so it can not be accomplished for our data .
however , as we describe below , all of these sources have well - calibrated @xmath19 magnitudes available from ukidss .
we choose instead to defer to those measurements in determining colors ( section 3.2 ) .
finally , all candidate companions with separations @xmath54.2 have spatially resolved counterparts in the ukidss galactic cluster survey @xcite , and many of the widest companions also have optical counterparts in the usno - b1.0 digitization of the palomar observatory sky survey @xcite .
as we describe further in section 3.2 , we have used those observations to identify most of these candidate companions as background stars .
the rest require multi - epoch astrometric monitoring to determine whether they are associated . around three targets in the k08 survey we found faint visual candidate companions with projected separations of 23 and brightnesses of @xmath201516 : gsc 06214 - 00210 , 1rxs j160929.1 - 210524 , and 1rxs j160703.4 - 203634 , whose basic properties are listed in table [ tabprops ] .
we re - observed these objects with the telescope - instrument - filter combinations as detailed in table 2 .
the expected surface density of unassociated background stars at these brightnesses is relatively low , but since upper scorpius is at moderate galactic latitude and is projected over the background milky way bulge , this probability was not sufficiently low that we could assume they were bound ulmc companions .
follow - up observations were therefore required to confirm common proper - motion . the density of background stars at @xmath21 is approximately 0.002 per square arcsecond ( kraus & hillenbrand 2010 , in prep . ) , which gives a 86% chance of a chance alignment within 2.5for at least one star in our sample , but only a 30% chance of all 3 candidates being chance alignments .
it is impossible to use proper motions to rule out chance alignments with other association members because the internal velocity dispersion of upper sco is only @xmath01kms@xmath22 ( @xmath01.5masyr@xmath22 , @xcite ) , meaning that they would be comoving even if they were only seen in chance alignment .
however , the surface density of young stars in upper sco is no more than 100deg@xmath23 @xcite or @xmath24arcsec@xmath23 , meaning that the probability of a chance alignment is negligible .
the proper motion of the upper scorpius association is ( -11.5,-23.5)masyr@xmath22 in equatorial coordinates , determined from the members listed in @xcite , and using updated proper motions from @xcite .
this is relatively small when compared to nearby moving groups , so an astrometric accuracy of order 2mas or 0.1% at 2is required in order to clearly determine whether a companion is co - moving or not on a 1 year time baseline . the nirc2 camera has been shown to have a stability better than this in the precise galactic center astrometric work of @xcite .
@xcite showed that the palmao adaptive optics system @xcite and the pharo near infrared ( nir ) camera are stable in distortion to @xmath0100@xmath25as over several months .
however , palmao underwent several upgrades between these common proper motion confirmation observations , including preliminary work for a reconfiguration of the output beam path to accommodate new science instruments .
we searched for palmao distortion solution changes during this period using observations of the core of the m5 globular cluster from four nights ( 28/5/2007 ; 29/5/2007 ; 17/07/2008 and 17/08/2008 ) .
the observations consisted of 100 co - added 1.4 second exposures at each epoch , were taken in the 25 mas pharo plate scale , used the br-@xmath26 narrow - band filter to avoid differential chromatic refraction , and were timed to be observed at very similar airmasses and hour angles .
the same ao guide star was used in each case , and care was taken to align each epoch to the same pointing within an arcsecond and to keep all other ao and camera parameters consistent .
we fitted 2d gaussians to derive the positions of 34 stars covering the 25``@xmath2725 '' field in each dataset ; we also checked the fitted positions using sextractor @xcite and found very similar results .
we first used a simple model to match the stellar positions between epochs , allowing for a change in pointing position , an arbitrary rotation , and a separate scaling in the x and y axes .
there was no detectable change in rotation or scale within 2007 or within 2008 , but between those years the field rotation changed by 0.18 degrees , while the x and y plate scales changed by -0.5 % and + 0.2% respectively .
however , the simple rotation and plate scale change model leaves 3 - 5mas residuals in matching the 2007 and 2008 measured positions .
the residuals are reduced to @xmath01mas using a general 4-parameter linear transformation , where the additional parameter represents the plate scale changes being in arbitrary orthogonal axes , or alternatively the addition of a shear term . however , some statistically significant higher - order distortion is still apparent after application of this solution .
it appears that long - term precision astrometry using adaptive optics requires careful attention to changes in the detailed distortion solutions ( beyond simple rotation and scale ) ; this is particularly important as , like palmao , the optics of many adaptive optics systems are regularly upgraded .
therefore , wherever at least two nirc2 epochs are available , we only consider the astrometry with nirc2 for our proper motion measurements .
all followup nirc2 data were acquired using the narrow camera with a @xmath010mas / pixel image scale and the @xmath28 filter , which we found did not quite saturate for these targets using the shortest exposure time with a minimum number of fowler samples .
the follow - up nirc2 data were reduced using a custom pipeline written in idl . after standard image - processing tasks ( background subtraction , flat - fielding and bad pixel removal )
were completed , the distortion in the image plane was then removed by the program nirc2dewarp available from the nirc2 camera home page , using the updated distortion solution of b.cameron ( personal communication , 2007 ) .
finally , aperture - photometry and centroids were computed with the imexam function of atv , version 2.0b4 .
selected data sets were also analysed with the daophot function of iraf , giving results consistent well within the error bars . for centroiding ,
a centering box size of 5 pixels was used , with the exception of one single observation in l - band , where a box size of 9 pixels was used .
the aperture - radius for photometry was 5 pixels for the kp filter , but 7 pixels for the j and h filters due to low strehls and dispersion , and 10 pixels in l - band due to the larger diffraction - limited core .
the sky annulus was always set at 10 to 20 pixels .
note that we did not attempt to calibrate absolute photometry , and have only computed the relative photometry between primary and secondary .
the following subsections describe the multi - epoch astrometry obtained on each of the three targets in table [ tabprops ] , in order to confirm proper motion .
two clearly associated objects are discussed , as well as two objects identified as background stars .
lccccrr gsc 06214 - 00210 & 16 21 54.67&-20 43 09.1&m1&0.60&11.6&9.15 + 1rxs j160929.1 - 210524&16 09 30.30&-21 04 58.9&k7/m0&0.68 - 0.77&12.1&8.92 + 1rxs j160703.4 - 203634&16 07 03.56&-20 36 26.5&m0+?&0.68 +
0.59&11.3&8.10 + [ tabprops ] gsc 06214 - 00210 is a young m1 star near the eastern edge of upper sco .
it was originally identified as a candidate young star by preibisch et al .
( 1998 ) based on its x - ray emission , then confirmed to have strong lithium absorption ( @xmath29 ) and h@xmath30 emission ( @xmath31 ) as compared to stars of equivalent spectral type in several young clusters .
it was also found to have a proper motion @xmath32 mas yr@xmath22 @xcite , consistent with the motion of upper sco . given its m1 spectral type , the 5 myr mass - temperature relations of @xcite predict a mass of 0.60@xmath33 . note that this mass must be taken with caution , as an uncertainty of 1 subclass in spectral type
should correspond to an uncertainty of @xmath00.1 @xmath33 in mass .
one should keep in mind that the models themselves carry an unknown uncertainty since the mass - luminosity and mass - temperature relations of young stars are almost completely uncalibrated for @xmath31 @xmath33 ( e.g. @xcite ) .
an example image of gsc 06214 - 00210b used for astrometry is shown in figure [ figimage ] .
our photometric and astrometric observations of this companion are given in table [ tabp98 - 70 ] and figure [ figp70 ] .
the candidate companion showed a relative motion of 7@xmath344 mas with respect to gsc 06214 - 00210 over 2.9 years .
this is much less than the @xmath070 mas motion expected if gsc 06214 - 00210b were a background star , and demonstrates that it is most likely physically associated with gsc 06214 - 00210 .
the apparent relative proper motion of [email protected]@xmath22 corresponds to [email protected] kms@xmath22 assuming a 7 mas parallax for upper scorpius , which is roughly equal to the circular orbital velocity of 1.7kms@xmath22 expected for a @xmath0300 au orbit . as can be seen from the larger contrast in bluer filters ( cf .
tabp98 - 70 ] ) , gsc 06214 - 00210b is quite red compared to its primary star .
the primary has an observed 2mass color of @xmath35 , and most m1 stars have a typical @xmath36 color of @[email protected] @xcite , so the inferred colors for the companion are @xmath37 and @xmath38 .
the mean distance for upper sco is 145 pc ( de zeeuw et al .
1999 ) , and a @xmath39 degree spread on the sky of the densest region ( where all our candidates belong ) corresponds to a @xmath014pc error on the distance of any individual star .
this gives a a distance modulus of @xmath40 , so given the 2mass magnitudes of the primary , the absolute magnitudes of the companion are @xmath41 , @xmath42 , @xmath43 , and @xmath44 .
the interstellar extinction toward upper sco is negligible ( @xmath45 or @xmath46 ) , so no extinction corrections should be required .
the observed @xmath47 color of 1.3 is consistent with the m8-l4 spectral type range of field dwarfs from @xcite .
the k - l color of 1.05 , however , would clearly place the star at the red end of this range , at l3-l4 .
the absolute k - magnitude of 9.1 is then about 2 magnitudes brighter than for corresponding field objects , providing clear evidence of a larger radius and young age ; this height above the main sequence is similar to that observed for other late - type members of upper sco @xcite and tw hya @xcite . the colors and spectral type of gsc 06214 - 00210b are not expected to match field dwarfs due to the low surface gravity .
@xcite find that j - k and k - l colors are significantly redder for low gravity objects at a given spectral type .
although gsc 06214 - 00210b is a low - mass object , @xcite focusses on redder object at j - k colors of @xmath02.0 , and it is nt clear that their result should hold for bluer objects like gsc 06214 - 00210b , with j - k=1.3 . in figure [ figcolorcolor ]
we show the nir the colors of gsc 06214 - 00210b with respect to field dwarfs and giants ( which bracket gsc 06214 - 00210b in gravity ) , where it appears to have a small but significant k - l excess .
the red k - l color could be due to a disk , as @xmath51/3 of young brown dwarfs retain a disk for @xmath55 myr ( e.g. @xcite ) , with a trend of increasing disk lifetime with decreasing mass .
spatially resolved spectroscopy of the system would be required in order to determine the spectral type more accurately , and to search for signs of accretion onto the secondary . given the observed brightness in the @xmath48 filters , a comparison to the 5 myr dusty models @xcite et al .
2000 ) suggests that the mass of the companion is @xmath01215 @xmath9 , with bluer filters suggesting slightly lower masses than redder filters .
the cond models @xcite never predict sufficiently red @xmath47 colors to match our observations , and indeed are not appropriate for objects with @xmath49k .
this trend qualitatively matches the observed trend for young low - mass objects to be redder than older field counterparts of similar spectral type .
if the comparison was based only on observed j - k color , then the dusty models would predict masses of 1012 @xmath9 . as we discussed above
, the @xmath50 photometry could have an excess from a circumstellar disk , so we suggest that the @xmath50 magnitude should not be used directly in the mass estimate . llrrr + 2454258.0 &
kp & [email protected] & [email protected] & [email protected] + 2454634.8 & kp & [email protected] & [email protected] & [email protected] + 2454634.8 & j & [email protected] & [email protected] & [email protected] + 2454982.9 & kp & [email protected] & [email protected] & [email protected] + 2455313.1 & kp & [email protected] & [email protected] & [email protected] + 2455313.1 & h & [email protected] & [email protected] & [email protected] + 2455313.1 & l & - & - & [email protected] + + 2454634.8 & kp & [email protected] & [email protected] & [email protected] + 2454982.9 & kp & [email protected] & [email protected] & [email protected] + + 2454250.8 & ks & 4261@xmath3414 & [email protected] & [email protected] + 2454982.9 & kp & 4215@xmath345 & [email protected] & [email protected] + + 2454634.8 & kp & [email protected] & [email protected] & [email protected] + 2454634.8 & j & [email protected] & [email protected] & [email protected] + 2454982.9 & kp & [email protected] & [email protected] & - + [ tabp98 - 70 ] 1rxs j160929.1 - 210524 is a young m0 star located near the center of upper sco . like gsc 06214 - 00210 ,
it was first identified as a likely upper sco member by preibisch et al .
( 1998 ) and exhibits both strong lithium absorption ( @xmath51 ) and h@xmath30 emission ( @xmath52 ) .
the proper motion reported by ucac3 @xcite for this object is ( -11.2,-21.9 ) @xmath341.5 mas yr@xmath22 , which is also consistent with the value for upper sco .
the spectral type reported by preibisch et al .
( 1998 ) was m0 , but , as was described by @xcite , newer measurements by d. c. nguyen suggest a more likely spectral type of k7 .
the inferred mass would be 0.68 - 0.77 @xmath33 for the two estimates , with uncertainties similar to that for gsc 06214 - 00210 .
our observations for 1rxs j160929.1 - 210524b are listed in table [ tabp98 - 70 ] and plotted in figure [ figp50 ] .
the companion was first detected in palomar images as early as may 2007 , but as the astrometric performance of the pharo camera was unverified on large timescales ( cf .
section [ sectfollowup ] ) , we were not able to convincingly demonstrate common proper motion until 2009 .
this object was independently detected by @xcite and further characterized by @xcite .
our astrometric results are consistent with those recently reported by @xcite at 2-@xmath18 in separation and at 1-@xmath18 in position angle , but have higher precision due to the accurate astrometric characterization of nirc2 . based on a spectrum indicating a late spectral type and low gravity , these authors argued that this faint companion was a young object and therefore likely a physical companion .
since that time , we have continued monitoring this object ( cf . table [
tabp98 - 70 ] , and can now confirm that 1rxs j160929.1 - 210524b is a physical companion of 1rxs j160929.1 - 210524 .
the apparent proper motion of [email protected]@xmath22 , corresponding to [email protected]@xmath22 .
this is consistent with the @xmath01.7kms@xmath22 orbital motion expected , especially considering possible projection effects .
since @xcite already reported nir colors , we only took @xmath19-band observations .
these measurements are consistent with theirs , and indicate an absolute magnitude of @xmath53 .
the predicted mass from the dusty and cond models is @xmath08 @xmath9 from the @xmath19 band photometry or @xmath07 @xmath9 from @xmath54 band photometry , though again , these estimates are completely uncalibrated by observations .
given the assumptions listed above for the distance of upper sco , the projected separation is @xmath0320 au , similar to the value for gsc 06214 - 00210 system .
unassociated background stars typically are identified based on multi - epoch astrometric monitoring ( which indicates that they are not comoving ) or multi - wavelength observations to measure colors ( which indicate that they do not fall along the same color - magnitude sequence ) .
as we describe in the next several subsections , several of the closer companions will require astrometric monitoring with high - resolution imaging to confirm or disprove their association , but the wider companions can be identified and rejected based on archival photometry from seeing - limited all - sky surveys .
in addition to the companion listed as companion b in table [ tabp98 - 70 ] , an additional wider companion candidate , listed as companion c in table [ tabp98 - 70 ] was found around 1rxs j160929.1 - 210524 in both the 2007 palomar images ( see section [ sectwider ] ) and the 2009 keck images .
this companion candidate was also reported by @xcite . although the palomar astrometry had significant errors , the time baseline of 2 years was sufficient to clearly show that this object was a background star , as shown in figure [ figp50c ] .
1rxs j160703.4 - 203634 is a young , close binary system which is also located near the center of upper sco .
it was originally identified in a survey for new sco - cen members by @xcite , using methods very similar to the survey by preibisch et al .
( 1998 ) : targets were identified as potential x - ray emitters , then confirmed to have strong lithium for their age . however , the survey by kunkel was never published in the refereed literature , so the equivalent widths of relevant spectral lines are not available .
we discovered its multiplicity during our direct imaging and aperture masking survey ( k08 ) , finding that it had a companion with projected separation of 184 mas ( @xmath027 au ) and flux ratio @xmath55 ( mass ratio @xmath56 ) .
the same observations also revealed a faint candidate companion at a projected separation of @xmath02.1 . as we summarize in table [
tabp98 - 70 ] and figure [ figk99 ] , multi - epoch astrometry for the candidate companion shows that it is not comoving with 1rxs j160703.4 - 203634 , but instead appears to be nearly stationary , as would be expected for a background star . the relative motion of the primary with respect to the companion is ( 7@xmath342,-26@xmath342)masyr@xmath22 , consistent with the proper motion of upper scorpius .
the contrast with respect to the brighter member of the close binary pair is @xmath57 and @xmath58 , indicating that the companion has a color of @xmath59 . given that the total extinction along this line of sight is only @xmath601.2 or @xmath610.2 ( schlegel et al .
1998 ) , the companion must be intrinsically cool ( @xmath620.7 or spectral type @xmath5k5 ; kraus & hillenbrand 2007 ) . combined with its nonmotion ( which indicates it is likely to be quite distant ) , we therefore conclude that the candidate companion is likely a background k or early m giant , perhaps located in the milky way bulge .
multi - epoch observing campaigns can be observationally expensive , so where possible , it is best to use archival data to rule out possible companions .
this is sometimes impossible since many candidate companions of interest can only be distinguished from their candidate primary using high - resolution imaging techniques .
however , wider companions can often be resolved in seeing - limited data , especially for new surveys that have very good spatial resolution ( i.e. , @xmath30.5 for ukidss images ) . of the 25 companions with separations of @xmath53.5 ,
21 were detected in both the @xmath63 and @xmath19 filters by ukidss , so we can use their @xmath64 colors to determine whether they might be associated .
ukidss observations of known low - mass members of upper sco by lodieu et al .
( 2007 ) show that most members fainter than @xmath2013 ( i.e. with spectral type @xmath5m7 ) have colors of @xmath65 ; this limit is consistent with the typical @xmath64 colors of @xmath5m7 field dwarfs as compiled in kraus & hillenbrand ( 2007 ) .
as we show in table 3 , none of the candidate companions with both @xmath63 and @xmath19 magnitudes meet this criterion , so we identify all of them to be unassociated field stars , most likely in the distant background behind upper sco .
in addition , 17 of the wider candidate companions have counterparts visible in the usno - b1.0 digitization of the palomar observatory sky survey @xcite .
most are not present in the usno - b1.0 source catalog since its source identification algorithm was extremely conservative in identifying faint neighbors to bright stars .
however , these sources can be manually identified by visual inspection . in table 3
, we list the bluest plate ( b or r ) at which each of the companions was visible . as we noted above , any true companions should have spectral types of @xmath5m7 , so our compilation of field dwarf colors ( kraus & hillenbrand 2007 ; kraus & hillenbrand , in prep ) suggests that the expected colors for true companions are @xmath66 and @xmath67 .
the detection limits of the poss survey were @xmath6821 and @xmath6920 ,
so all 17 sources with counterparts in the b or r plates must be background stars that are bluer than these limits .
in all cases , these identifications agree with the ukidss identification . three wider companions in table [ tblwider ]
can not be eliminated as upper scorpius members due to insufficient photometry or astrometry .
these are the companions to gsc 06793 - 00994 , gsc 06794 - 00156 and scopms 015 .
the closest of these systems is scopms 015 , with a companion just outside of 500au .
this small incompleteness defines the outer limit of our survey until additional follow - up observations can be obtained , though some azimuthal uncertainty remains at separations of @xmath6500 au due to image boundaries ( section [ sectdectlimits ] ) .
lrrrcrrr rxj1603.6 - 2245&10959@xmath3434&[email protected]&[email protected]&r2&[email protected]&[email protected]&[email protected] + rxj1603.9 - 2031a&9484@xmath3429&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + rxj1606.2 - 2036&6820@xmath3422&[email protected]&[email protected]& .. &[email protected]&[email protected]&[email protected] + rxj1607.0 - 2036&11939@xmath3436&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + usco-160517.9 - 202420&7638@xmath344&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + usco-160801.4 - 202741&7175@xmath3424&[email protected]&[email protected]& .. &[email protected]&[email protected]&[email protected] + usco-160801.4 - 202741&10666@xmath3432&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + usco-160900.7 - 190852&6850@xmath3427&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc 06205 - 00954&15398@xmath3447&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc 06209 - 01501&8994@xmath3427&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc
06213 - 01358&4261@xmath3414&[email protected]&[email protected] & .. & .. & .. & .. + gsc 06793 - 00797&11086@xmath3433&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc 06793 - 00797&12330@xmath3438&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc
06793 - 00994&5462@xmath3417&[email protected]&[email protected]& .. & .. &[email protected] & .. + gsc 06794 - 00480&11820@xmath3436&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc 06214 - 00210&12941@xmath3439&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + gsc
06794 - 00537&15942@xmath3449&[email protected]&[email protected]&
.. &[email protected]&[email protected]&[email protected] + gsc 06794 - 00537&5214@xmath3416&[email protected]&[email protected]& .. &[email protected]&[email protected]&[email protected] + gsc 06794 - 00156&5973@xmath3418&[email protected]&[email protected]& .. & .. &[email protected] & .. + scopms 015&16954@xmath3451&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + scopms 015&3538@xmath3411&[email protected]&[email protected] & .. & .. & .. & .. + scopms 015&7485@xmath3423&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + scopms 015&12482@xmath3438&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + scopms 045&7063@xmath3421&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + scopms 045&11286@xmath3434&[email protected]&[email protected]&b2&[email protected]&[email protected]&[email protected] + [ tblwider ] in order to infer the properties of the distribution of wide ulmcs , it is essential not only to establish the existance of a small number of physical companions , but also to determine the magnitude limit as a function of separation for possible companions not confidently identified in the observations .
this has not been general practice in previously reported wide ulmcs , in particular not for several of the @xmath340m@xmath8 companions in our separation and primary mass range : gq lup b , ct cha b and 1rxs j160929.1 - 210524 b. in this section , we will describe the detection limits for all stars in our sample .
as we described in section [ sectdisco ] , we measured our detection limits in a method that accounts for both spurious detections from speckle noise ( at small separations ) and the sky background limit ( at large separations ) .
these limits could in principle be used as input for monte carlo or bayesian techniques to study the underlying population , though as we discuss in the section [ sectfreq ] , using the results from our survey alone could yield a biased measurement .
table [ tablimits ] lists and figure [ figlimits ] shows the companion detection limits for each star in the sample .
the last column of table [ tablimits ] also gives the maximum separation in arcseconds where we surveyed all position angles for companions .
the few small ( @xmath33 ) values of @xmath70 in this column are due to quick image sets taken in the camera sub - array mode we used for aperture - masking interferometry .
the limits at separations smaller than 1 are set by the separation - dependent speckle noise and extended psf halo of the primary star , while the wider limits at separation greater than 1.5are constant and result from the sky background ( for broadband @xmath19 observations ) or read noise ( for narrowband @xmath13 observations ) . for each star
, we list the primary @xmath19 magnitude ( from 2mass , with the flux from any close binary companions subtracted ) , the detection limit in @xmath71 at a range of angular separation , and the corresponding detection limit in @xmath9 at the corresponding projected orbital distances .
the mass detection limits were derived from the k magnitudes of the 5 myr dusty isochrone @xcite .
lrrrrrrrcccccccc gsc 06205 - 00954&12.9&13.4&13.4&14.1&15.0&16.4&17.8&42 & 28 & 28 & 21 & 14&6.9&4.5&10.4 + gsc 06208 - 00834&12.6&13.5&13.6&14.0&14.8&16.4&17.5&57 & 27 & 26 & 22 & 16&6.8&4.7&10.7 + gsc 06209 - 01501&12.3&13.3&13.1&13.7&14.8&16.1&17.7&74 & 30 & 33 & 25 & 16&7.6&4.5&10.4 + gsc 06213 - 00194&12.1&12.7&13.3&13.6&14.0&15.7&17.1&78 & 52 & 30 & 26 & 22&11&5.4&10.2 + gsc 06213 - 01358&12.9&13.6&13.8&14.2&15.2&16.5&17.9&39 & 26 & 23 & 20 & 14&6.5&4.4&10.4 + gsc 06214 - 00210&12.5&13.3&13.8&14.6&14.8&16.0&17.4&62 & 29 & 24 & 17 & 16&8.4&4.8&9.4 + gsc 06214 - 02384&12.3&12.9&13.5&13.9&14.4&15.8&17.5&76 & 41 & 27 & 23 & 18&9.9&4.8&10.2 + gsc 06764 - 01305&12.4&13.1&13.6&14.1&14.8&16.2&16.9&71 & 33 & 27 & 21 & 16&7.4&5.7&11.1 + gsc 06793 - 00797&11.7&12.9&13.2&13.8&14.4&16.0&17.3&97 & 42 & 31 & 24 & 19&8.3&5.0&10.2 + gsc 06793 - 00994&12.1&12.8&13.4&13.5&14.1&15.7&17.2&78 & 43 & 29 & 28 & 21&10&5.2&10.2 + gsc 06794 - 00156 & 9.8&10.7&11.4&12.1&12.9&14.6&17.2&413&212&130&78 & 41&17&5.1&10.2 + gsc
06794 - 00480&12.0&12.7&13.1&13.8&13.9&16.0&17.0&81 & 53 & 33 & 24 & 23&8&5.5&10.2 + gsc 06794 - 00537&11.9&12.7&13.1&13.6&14.0&15.7&17.7&89 & 54 & 34 & 26 & 22&10&4.6&10.2 + rxj1550.0 - 2312 & 13.7&14.4&15.1&16.8&17.4&17.2&17.1&25 & 19 & 14 & 5.8&4.9&5.1&5.4&2.1 + rxj1550.9 - 2534 & 11.4&12.8&13.8&15.6&16.8&17.1&17.3&129&44 & 24 & 11 & 5.9&5.4&4.9&1.3 + rxj1551.1 - 2402 & 14.3&15.5&15.8&17.0&17.1&17.1&17.2&19 & 12 & 10 & 5.5&5.3&5.3&5.2&1.5 + rxj1557.8 - 2305 & 13.7&15.0&15.5&16.9&17.3&17.2&17.1&25 & 15 & 12 & 5.7&4.9&5.1&5.4&3.7 + rxj1558.1 - 2405 & 11.4&12.5&13.2&14.1&14.6&16.1&17.0&126&64 & 32 & 21 & 17&7.7&5.6&8.9 + rxj1558.2 - 2328 & 11.4&12.1&12.7&13.3&13.9&15.4&17.1&128&78 & 54 & 29 & 23&12&5.2&10.2 + rxj1600.7 - 2127 & 12.5&13.6&14.1&14.2&15.0&16.3&17.6&61 & 26 & 21 & 20 & 15&7.1&4.6&10.6 + rxj1601.1 - 2113 & 12.7&13.4&13.5&13.7&14.8&16.5&17.4&56 & 28 & 27 & 25 & 16&6.6&4.8&9.6 + rxj1601.9 - 2008 & 11.7&12.2&12.3&12.9&13.5&15.6&17.6&101&76 & 72 & 39 & 28&11&4.6&11 + rxj1602.0 - 2221 & 13.1&14.4&15.1&16.8&16.9&17.0&17.1&35 & 19 & 14 & 5.8&5.7&5.5&5.3&3.7 + rxj1602.8 - 2401a&11.1&11.8&12.4&14.1&15.0&16.0&17.0&156&95 & 71 & 20 & 15&8.4&5.6&3.6 + rxj1602.8 - 2401b&10.1&10.6&11.3&12.4&13.3&14.7&15.1&332&226&131&67 & 30&16&14.0&10.2 + rxj1603.6 - 2245 & 11.9&12.7&13.4&13.8&14.4&15.8&17.4&89 & 54 & 28 & 24 & 19&9.3&4.8&10.2 + rxj1603.9 - 2031a&11.9&12.7&13.4&13.6&14.5&16.0&17.6&88 & 54 & 29 & 27 & 18&8.4&4.6&10.1 + rxj1604.3 - 2130 & 11.9&12.3&13.0&13.7&14.3&15.9&17.1&83 & 73 & 37 & 25 & 19&8.7&5.2&8.9 + rxj1606.2 - 2036 & 10.4&11.1&11.7&12.8&13.5&15.1&15.4&271&164&96 & 45 & 27&14&12.0&11.1 + rxj1607.0 - 2036 & 9.1&12.4&13.3&13.9&14.5&15.6&17.9&648&69&30&23&18&11&4.3&8.7 + scopms015&11.3 & 13.4&13.8&14.1&15.1&16.9&17.9&140&28&24&20&14&5.7&4.3&10.4 + scopms017&13.9 & 14.7&15.0&16.9&17.6&17.4&17.2&23&17&15&5.7&4.6&4.9&5.2&1.5 + scopms019&11.3 & 12.1&12.5&13.5&13.9&15.5&16.6&133&79&65&27&23&11&6.2&8.8 + scopms022&14.3 & 15.5&15.5&17.4&17.4&17.3&17.3&19&12&12&4.9&4.9&4.9&5.0&1.4 + scopms027&11.8 & 12.6&12.7&12.9&13.5&15.3&16.4&89&57&49&42&28&13&6.9&3.0 + scopms028&13.1 & 15.0&15.9&17.1&17.2&17.0&16.9&34&15&9.3&5.3&5.1&5.4&5.7&3.7 + scopms044&11.8 & 12.3&12.7&13.0&13.4&15.5&17.7&91&72&53&36&28&12&4.5&10.3 + scopms045&12.8 & 13.4&13.8&14.2&14.6&15.8&18.0&48&28&24&20&17&9.6&4.2&10.7 + usco-160341.8 - 200557&14.0&15.4&15.7&17.3&17.4&17.4&17.3&22 & 12 & 10 & 5 & 4.8&4.9&5.0&2.2 + usco-160643.8 - 190805&10.1&10.8&11.6&13.2&13.8&15.6&15.8&336&203&110&30&24&11&9.4&8.5 + usco-160707.7 - 192715&14.6&16.1&16.6&17.9&17.9&17.7&17.6&17 & 7.6&6.2&4.4&4.3&4.5&4.7&1.5 + usco-160801.4 - 202741&11.1&11.9&12.5&13.7&14.1&15.8&16.1&158&86 & 64 & 25&20&10&7.8&8.5 + usco-160823.2 - 193001&14.4&15.5&15.7&17.3&17.4&17.3&17.3&18 & 11 & 11 & 5&4.9&5.0&5.0&3.7 + usco-160825.1 - 201224&14.3&15.8&16.3&17.3&17.5&17.5&17.5&19 & 10 & 7 & 4.9&4.8&4.8&4.8&1.5 + usco-160900.7 - 190852&10.8&11.5&12.2&13.6&14.1&15.6&16.0&196&115&76 & 26&21&11&8.3&6.0 + usco-160916.8 - 183522&14.3&15.8&16.0&17.4&17.7&17.6&17.4&19 & 9.6&8.1&4.9&4.5&4.7&4.9&2.1 + usco-160954.4 - 190654&14.5&15.5&16.0&17.4&17.4&17.5&17.5&18 & 12 & 8.4&4.9&4.9&4.8&4.7&2.2 + usco-161031.9 - 191305&10.8&11.5&12.2&13.3&13.9&15.5&16.1&200&114&77 & 29&23&12&7.7&8.5 + usco-161347.5 - 183459&13.4&13.5&13.5&13.4&13.3&13.2&13.1&28 & 27 & 27 & 29&29&31&34&2.1 + [ tablimits ]
the most basic step in assessing the ability of current planet formation theories ( see section 1 ) to explain the widely separated and relatively massive candidate planetary companions that are observed is to measure their frequency .
most of the 49 stars in our sample have relatively uniform detection limits ( @xmath046 @xmath9 at @xmath5300au ) , so a naive estimate of the frequency is approximately 2/49=4.1@xmath72% .
here we have quoted the most likely frequency , and the bayesian 68% confidence interval on the frequency with a prior distribution where all frequencies are equally likely .
the detection limits are not completely uniform , so the exact frequency will depend on more sophisticated analysis using monte carlo ( k08 ) or bayesian ( @xcite ; kraus and hillenbrand , in prep . )
techniques , and ultimately should depend on the separation and mass distributions of ulmcs .
there is also a more fundamental issue that must be considered : this survey is not the first to be sensitive to the presence of ulmcs , and a full treatment should consider all surveys that have properly reported null detections and detection limits ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite reported 1 detection ( ab pic ) from 30 - 40 young solar - type targets surveyed ( depending on the definitions of `` young '' and `` solar - type '' ) .
indeed , the full sample of past null detections is larger than our observed sample , suggesting that the true frequency could be lower by up to a factor of @xmath52 , and that our survey had good fortune to discover two new companions . alternatively , including the companions to ct cha and possibly gq lup ( likely more massive than 20@xmath2 ) may increase the true frequency of wide ulmcs , if only the details of the survey samples in which these companions were discovered were known .
a large census of the literature is beyond the scope of a discovery paper , and the results of this analysis will be reported in a companion paper ( kraus et al .
, in prep . )
the population of directly imaged wide ( @xmath540au ) ulmcs poses a significant challenge to planet formation models .
existing models of solar - system scale planet formation ( i.e. , via core accretion or gravitational instability in a class ii disk ; @xcite ) have not been succesful at forming planets at the orbital radii they are observed here ( @xmath5200 au ) . it is also difficult to form wide ulmcs like a binary ( i.e. via fragmentation of the freefalling protostellar core or fragmentation in the massive protostellar disk ) without subsequently accreting sufficient mass to become a stellar or brown dwarf companion .
however , one of these mechanisms must occur . complicating this picture
is the possibility , and indeed likelihood , of multiple planet scattering in some formation scenarios .
therefore , we can not answer the question of formation mechanism without examining the population of ulmcs in separation and mass space alongside more well studied classes of companions .
the core - accretion mechanism for planetary formation only operates close to the host star ( @xmath05au ) , where protoplanetary disk densities are high enough to enable dust and ice to coagulate into protoplanets .
therefore , a wide ulmc could only originate from core - accretion if it came from a scattering event . in any scattering event , low - mass companions are preferentially scattered outwards , so a core - accretion and scattering origin for wide ulmcs would require an even greater population of close , higher - mass objects . as brown dwarfs are well known to be rare less than @xmath05au from their host stars ( the brown - dwarf desert
* ) , this scenario is very unlikely .
another plausible formation scenario for wide ulmcs appears to be formation like stars from the fragmentation of protostellar clouds and massive circumstellar disks ( e.g. @xcite ) . in existing models , it is difficult to form ulmcs because fragments subsequently accrete mass , and the only fragments that remain near the opacity limit for fragmentation are those that are ejected @xcite . in an extreme case , the fragmentation of each core might be expected to be independent and companions would be expected to follow an initial - mass function ( i.e. random pairing ) .
this is now well known to be incorrect , with an approximately linearly flat distribution of companion masses for solar type primaries and @xmath010@xmath73au separations ( @xcite ; @xcite ; kraus et al 2010 in prep . ) . as young solar - type stars have stellar companion fractions in the range 1222% ( @xcite ; k08 ; kraus et
al , 2010 , submitted ) per decade of separation , this would mean that 0.2 - 0.3% of solar - type stars should have @xmath050 - 500au , @xmath740.006 to 0.02 companions ( i.e. 620@xmath2 for solar - type stars .
we see a clear surplus to this model , strongly suggesting that wide ulmcs follow a different formation path to stars . if wide ulmcs were to form via fragmentation of a circumstellar disk when the primary is at the class ii stage ( disk masses 0.001 - 0.1m@xmath75 , e.g. @xcite ) it would be much easier to form wide companions than with core accretion .
although most observed disks have masses of @xmath05m@xmath8 , a few disks around solar - type stars ( e.g. dl tau ) have large enough linear dimensions and mass ( up to @xmath01000au and 0.1@xmath76 , @xcite ) to fragment into observed ulmcs .
relatively little work has been done examining in detail how these large disks might fragment , with the bulk of disk - fragmentation literature discussing the possible formation of solar - system scale planets via fragmentation .
where models of large disks have been computed ( e.g. @xcite ) , it is clear that it is easier for them to become toomre - unstable and fragment than @xmath020au disks .
the remaining questions to be answered about this fragmentation include how many fragments are expected , and if the fragments can accrete enough of the disk mass to become objects like gsc 06214 - 00210 b and 1rxs j160929.1 - 210524 b. the question of where each formation mechanism operates is certainly not answered yet , but could be in the next few years .
radial velocity techniques will provide real constraints on the @xmath05 - 10au giant planet frequency as their time baselines increase , while direct imaging surveys will finish probing the wide ulmc frequency around young stars .
finally , high angular resolution techniques such as aperture - masking and high - efficiency coronography will fill much of the yet - unprobed @xmath05 - 30au regime of orbital separations which likely forms the boundary between the dominance of core - accretion and fragmentation processes .
m.i . would like to acknowledge support from the australian research council through an australian postdoctoral fellowship .
we would like to express thanks to undergraduate students alison hammond and matthew hill from the university of sydney who made a first - pass astrometric analysis of the data .
alk was suported by a sim science study and by nasa through hubble fellowship grant 51257.01 awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc . , for nasa , under contract nas 5 - 26555 .
some of these observations were obtained at the hale telescope at palomar observatory , as part of a collaborative agreement between the california institute of technology , jpl and cornell university .
some of the data presented herein were obtained at the w.m .
keck observatory , which is operated as a scientific partnership among the california institute of technology , the university of california and the national aeronautics and space administration .
the observatory was made possible by the generous financial support of the w.m .
keck foundation .
the authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community .
we are most fortunate to have the opportunity to conduct observations from this mountain . | at wide separations , planetary - mass and brown dwarf companions to solar type stars occupy a curious region of parameters space not obviously linked to binary star formation or solar - system scale planet formation .
these companions provide insight into the extreme case of companion formation ( either binary or planetary ) , and due to their relative ease of observation when compared to close companions , they offer a useful template for our expectations of more typical planets .
we present the results from an adaptive optics imaging survey for wide ( @xmath050 - 500au ) companions to solar type stars in upper scorpius .
we report one new discovery of a @xmath014m@xmath1 companion around gsc 06214 - 00210 , and confirm that the candidate planetary mass companion 1rxs j160929.1 - 210524 detected by @xcite is in fact co - moving with its primary star . in our survey , these two detections correspond to @xmath04% of solar type stars having companions in the 6 - 20@xmath2 mass and @xmath0200 - 500au separation range . this figure is higher than would be expected if brown dwarfs and planetary mass companions were drawn from an extrapolation of the binary mass function . finally , we discuss implications for the formation of these objects .
[ firstpage ] |
fusion frames , as a generalization of frames , are valuable tools to subdividing a frame system into smaller subsystems and combine locally data vectors .
the theory of fusion frames was systematically introduced in @xcite . since then
, many useful results about the theory and application of fusion frames have been obtained rapidly @xcite . in the context of signal transmission , fusion frames and their alternative duals
have important roles in reconstructing signals in terms of the frame elements .
the duals of fusion frames for experimental data transmission are investigated in @xcite .
but the problem that occurs is that the duality properties of fusion frames are not like discrete frames , such as , the duality property of fusion frames is not alternative and fusion riesz bases have more than one dual .
this paper deals with investigating such problems , which help us to obtain alternative dual fusion frames .
let @xmath0 be a separable hilbert space .
frame _ for @xmath0 is a sequence @xmath1 such that there are constants @xmath2 satisfying @xmath3 the constants @xmath4 and @xmath5 are called _ frame bounds_. if @xmath6 , we call @xmath7 a _ tight frame_. if the right - hand side of ( [ def frame ] ) holds , we say that @xmath7 is a _
bessel sequence_. given a frame @xmath7 , the _ frame operator _ is defined by @xmath8 a direct calculation yields @xmath9 hence , the series defining @xmath10 converges unconditionally for all @xmath11 and @xmath12 is a bounded , invertible , and self - adjoint operator .
hence , we obtain @xmath13 the possibility of representing every @xmath11 in this way is the main feature of a frame .
a sequence @xmath7 is bessel sequence if and only if the operator @xmath14 , which is called the _ synthesis operator _ , is well - defined and bounded . when @xmath7 is a frame , the synthesis operator t is well - defined , bounded and onto .
a sequence @xmath15 is called a _
dual _ for bessel sequence @xmath7 if @xmath16 every frame at least has a dual .
in fact , if @xmath7 is a frame , then ( [ frame decomposition ] ) implies that @xmath17 , which is a frame with bounds @xmath18 and @xmath19 , is a dual for @xmath7 ; it is called the _ canonical dual_. to see a general text in frame theory see @xcite .
let @xmath20 and @xmath21 be bessel sequences with synthesis operators @xmath22 and @xmath23 , respectively . then from ( [ def dual ] ) follows immediately that @xmath20 and @xmath21 are dual of each other if and only if @xmath24 ; in particular , they are frames . for more studies in the duality properties of frames
we refer to @xcite .
the following proposition describes a characterization of alternate dual frames .
@xcite [ ar ] 1 .
the dual frames of @xmath20 are precisely as @xmath25 , where @xmath26 is a bounded left inverse of @xmath27 and @xmath28 is the canonical orthonormal basis of @xmath29 2 .
there is a one to one correspondence between dual frames of @xmath20 and operators @xmath30 such that @xmath31 .
we now review preliminary results about fusion frames . throughout this paper
, @xmath32 denotes a countable index set and @xmath33 is the orthogonal projection onto a closed subspace @xmath34 of @xmath0 .
let @xmath35 be a family of closed subspaces of @xmath0 and @xmath36 be a family of weights , i.e. @xmath37 , @xmath38 .
then @xmath39 is a _ fusion frame _ for @xmath0 if there exist constants @xmath2 such that @xmath40 the constants @xmath4 and @xmath5 are called the _ fusion frame bounds_. if we only have the upper bound in ( [ def fusion ] ) we call @xmath39 a _ bessel fusion sequence_. a fusion frame is called _ tight _ , if @xmath4 and @xmath5 can be chosen to be equal , and _ parseval _ if @xmath41 . if @xmath42 for all @xmath38 , the collection @xmath39 is called _ @xmath43-uniform_. a fusion frame @xmath39 is said to be an _
orthonormal fusion basis _
if @xmath44 , and it is a _ riesz decomposition _ of @xmath0
if for every @xmath11 there is a unique choice of @xmath45 so that @xmath46 .
recall that for each sequence @xmath35 of closed subspaces in @xmath0 , the space @xmath47 with the inner product @xmath48 is a hilbert space . for a bessel fusion sequence @xmath39 for @xmath0 , the _ synthesis operator _
@xmath49 is defined by @xmath50 its adjoint operator @xmath51 which is called the _ analysis operator _ is given by @xmath52 recall that @xmath53 is a fusion frame if and only if the bounded operator @xmath54 is onto @xcite and its adjoint operator @xmath55 is ( possibly into ) isomorphism .
if @xmath39 is a fusion frame , the _ fusion frame operator _ @xmath56 is defined by @xmath57 is a bounded , invertible and positive operator and we have the following reconstruction formula @xcite @xmath58 the family @xmath59 , which is also a fusion frame , is called the _ canonical dual _ of @xmath39 and satisfies the following reconstruction formula @xcite @xmath60 let @xmath39 be a fusion frame by the frame operator @xmath61 .
a bessel fusion sequence @xmath62 is called a _ dual _ of @xmath39 if @xmath63 let @xmath35 be a family of closed subspaces of @xmath0 and @xmath36 be a family of weights , i.e. @xmath37 , @xmath38 .
we say that @xmath64 is a _ fusion riesz basis _ for @xmath0 if @xmath65 and there exist
constants @xmath66 such that for each finite subset @xmath67 @xmath68 some characterizations of fusion riesz bases are given in the following theorem .
[ rie]@xcite let @xmath69 be a fusion frame for @xmath0 and @xmath70 be a basis for @xmath71 for each @xmath72 .
then the following conditions are equivalent .
+ ( 1 ) @xmath73 is a riesz decomposition of @xmath0 .
+ ( 2 ) the synthesis operator @xmath54 is one - to - one .
+ ( 3 ) the analysis operator @xmath55 is onto .
+ ( 4 ) @xmath74 is a fusion riesz basis for @xmath0 .
+ ( 5 ) @xmath75 is a riesz basis for @xmath0 .
[ har]@xcite let @xmath76 and @xmath77 be a closed subspace .
then we have @xmath78 this paper is organized as follows : in section 2 , we compare the duality properties of discrete and fusion frames and by presenting examples of fusion frames we show that some well - known results on discrete frames are not valid on fusion frames .
also we investigate the cases that these properties can satisfy on fusion frames . in section 3
, we investigate the relation between the duals of fusion frames , local frames and the associated discrete frames and we try to characterize the dual of fusion frames .
for a fusion frame @xmath39 and a bessel fusion sequence @xmath62 , we define @xmath79 by @xmath80 it is easy to see that @xmath81 is a linear operator and @xmath82 , its adjoint can be given by @xmath83 , for all @xmath84 .
now , the identity ( [ def : alt ] ) can be written in an operator form as follows .
[ dual by preframes ] let @xmath39 be a fusion frame .
a bessel fusion frame @xmath62 is a dual of @xmath85 if and only if @xmath86 where @xmath54 and @xmath87 are the synthesis operators of @xmath35 and @xmath88 , respectively .
by lemma [ dual by preframes ] , we deduce that , unlike discrete frames , two fusion frames are not dual of each other in general . here
we present an example which confirms this statement .
[ s ] let @xmath89 .
consider @xmath90 @xmath91 and @xmath92 , for @xmath38 . also take @xmath93 @xmath94 and @xmath95 , for @xmath38
. then @xmath96 and @xmath97 are fusion frames for @xmath98 with frame operators @xmath99 and @xmath100 , respectively .
the following calculation shows that @xmath101 is an alternative dual of @xmath74 .
@xmath102\\&=&(a , b , c),\ \ \ ( a , b , c)\in\mathbb{r}^{3}.\end{aligned}\ ] ] but @xmath74 is not an alternative dual of @xmath101 . in fact
@xmath103\neq ( a , b , c).\end{aligned}\ ] ] now , it is natural to ask when two bessel fusion frames are dual of each other . to answer this question
assume that @xmath85 is also a dual fusion frame for @xmath104 or equivalently ( by lemma [ dual by preframes ] ) @xmath105 where @xmath106 is given by @xmath107 let @xmath73 be a fusion frame with a dual @xmath108 .
then the fusion frame @xmath35 is also a dual of @xmath88 if @xmath109 moreover the converse is hold if @xmath110 and @xmath111 are fusion riesz bases .
let @xmath88 be a dual of @xmath35 , then by using ( [ u.phi.t * ] ) and ( [ dual each other ] ) we obtain @xmath112 i.e. fusion frame @xmath35 is also a dual for @xmath88 .
+ for the proof of moreover part , since @xmath110 and @xmath111 are fusion riesz bases , by theorem [ rie ] , @xmath54 and @xmath113 are invertible .
so we deduce the proof by ( [ u.phi.t * ] ) and ( [ khan ] ) .
let @xmath114 and @xmath115 for each @xmath38 .
suppose that @xmath74 is a tight fusion frame for @xmath0 .
then @xmath74 is also a dual fusion frame of @xmath116 one of the important results in the duality of discrete frames is that every riesz basis has just a unique dual ( canonical dual ) and that dual is riesz basis as well .
but the following example shows that this property is not confirmed in fusion riesz bases .
[ dual riesz ] consider @xmath117 then @xmath118 is a fusion frame for @xmath98 with bounds @xmath119 and @xmath120 , and the frame operator @xmath121 it is not difficult to see that @xmath118 is a fusion riesz basis and its canonical dual can be given with @xmath122 to construct an alternate dual consider @xmath123 then @xmath124 is a fusion frame for @xmath98 .
moreover , if @xmath125 then @xmath126 hence , the fusion riesz basis @xmath118 has more than one dual and the second dual is not a fusion riesz basis .
let @xmath64 be a fusion frame for @xmath0 . by considering a frame for each subspace @xmath127 we can construct a discrete frame for @xmath0 .
we begin with the following key theorem .
[ relation fus.dis]@xcite for each @xmath38 let @xmath37 and let @xmath128 be a frame sequence in @xmath0 with the frame bounds @xmath129 and @xmath130 .
define @xmath131 for all @xmath38 and assume that @xmath132 then @xmath133 is a frame for @xmath0 if and only if @xmath64 is a fusion frame for @xmath0 .
our aim in this section is to study the relation between the duals of fusion frames , local frames and the associated discrete frames of @xmath0 . in particular , in the following theorem we investigate the relation between the duals of local frames of @xmath71 with the associated discrete frames of @xmath0 .
suppose that @xmath64 is a fusion frame for @xmath0 and @xmath143 is the frame operator of local frames @xmath144 for each @xmath38 .
now the question is whether the canonical dual of each frame @xmath144 is also a frame for the canonical dual of @xmath64 .
the following example shows that the answer is not true in general .
[ exam in r3 ] let @xmath145 consider @xmath146 @xmath147 and @xmath148 then by example 3.1 in @xcite , @xmath64 is a fusion frame for @xmath98 .
let @xmath149 and @xmath150 .
it is clear that @xmath151 is a frame for @xmath71 for each @xmath38 .
suppose that @xmath61 is the frame operator of @xmath152 and @xmath143 is the frame operator of @xmath151 for each @xmath38 .
a straightforward calculation shows that @xmath153 and the subspaces @xmath154 @xmath155 with the weights @xmath36 is the canonical dual of @xmath156 . moreover , if we take @xmath157 then @xmath158 is the canonical dual of @xmath151 for each @xmath38 .
however , @xmath158 is not a frame for @xmath159 for each @xmath38 .
the following example shows that there is no significant relation between the duals of fusion frames and their associated discrete frames , i.e. if @xmath160 is a dual of @xmath161 , then it is not necessary that their associated discrete frames be dual of each other .
let @xmath90 @xmath91 and @xmath93 @xmath162 then @xmath163 is a fusion frame for @xmath98 with an alternate dual @xmath164 .
+ consider @xmath165 and @xmath166 then @xmath167 and @xmath168 are frames for @xmath98 , but they are not dual of each other . in the rest of the paper
we try to characterize the duals of fusion frames .
we first discuss the riesz case .
let @xmath39 be a riesz decomposition of @xmath0 and @xmath62 be its dual .
associated to the canonical dual @xmath59 we can consider the operator @xmath181 given by @xmath182 applying ( [ u.phi.t * ] ) and theorem [ rie ] we conclude that @xmath183 , where @xmath184 is the synthesis operator of @xmath185 .
it follows easily that @xmath186 or equivalently @xmath187the following example shows that unfortunately , we can not characterize the duals of fusion frames by the duals of their associated discrete frames and the first part of proposition [ ar ] .
consider the fusion frame @xmath188 introduced in example [ exam in r3 ] . by theorem [ relation fus.dis ]
the sequence @xmath189 is frame for @xmath190 with the frame operator @xmath191 denote its canonical dual by @xmath192 . then @xmath193 consider @xmath194 @xmath195 and @xmath196 , then @xmath197 is a fusion frame for @xmath98 .
but @xmath197 is not an alternative dual of @xmath198 and vise - versa .
let @xmath73 be a fusion frame for @xmath0 and @xmath199 be a bessel sequence of normalized vectors such that @xmath200 . take @xmath201 where @xmath202 is the 1-dimensional subspace generated by @xmath203 .
then @xmath108 is a dual for @xmath73 .
first , it is not difficult to see that @xmath204 now by using 8.12 of @xcite and corollary 2.5 of @xcite we conclude that @xmath205 where @xmath5 and @xmath206 are the frame bounds of @xmath73 and @xmath199 , respectively .
hence , @xmath207 is a bessel fusion frame .
moreover , by lemma [ har ] we have @xmath208 the above theorem gives us a very simple method to construct duals of finite fusion frames . more precisely ,
let @xmath73 be a finite fusion frame for @xmath0 .
take @xmath209 where @xmath202 is a 1-dimensional subspace of @xmath210 . to illustrate this algorithm ,
let us consider the fusion frame @xmath35 in example [ dual riesz ] .
clearly @xmath211 @xmath212 @xmath213 therefore , we can introduce some duals : @xmath214 @xmath215 @xmath216 @xmath217 | fusion frames are valuable generalizations of discrete frames . most concepts of fusion frames are shared by discrete frames .
however , the dual setting is so complicated . in particular , unlike discrete frames , two fusion frames are not dual of each other in general . in this paper
, we investigate the structure of the duals of fusion frames and discuss the relation between the duals of fusion frames with their associated discrete frames . |
cognitive radio ( cr ) is an emerging wireless paradigm that enables wireless devices , called secondary users ( sus ) , to share the available spectrum whenever the licensed primary users ( pus ) are idle @xcite .
however , this dynamic spectrum sharing can introduce many security threats , one of which is the primary user emulation ( pue ) attack @xcite . in pue attacks ,
a malicious attacker emulates the pu signals to force legitimate sus to vacate the spectrum even though no real pus are currently present .
this type of attacks significantly decreases the total performance of legitimate sus and is detrimental to the overall cr system .
many defense and detection strategies have been proposed to prevent pue attacks in cr systems .
these strategies fall into the following main categories : authentication and encryption based such as in @xcite , localization - based @xcite , transmit power estimation techniques @xcite , and smart strategies based on supervised machine learning techniques such as in @xcite and @xcite or game theory @xcite . for pu authentication and encryption .
the authors in @xcite proposed to add helper nodes to the system that authenticate pus through link signatures , and hence inform sus with available spectrum . within the scope of localization - based techniques
, the authors in @xcite exploited the received signal strength ( rss ) distribution at different locations in the system for device fingerprinting .
however , both the authentication and the localization - based techniques will require additional nodes for authentication or location reference , which can increase system complexity .
localization - based techniques also assume the pus to be stationary with minimal channel variations , which is clearly not the case in practical crs in which users constantly change their locations . in @xcite ,
the authors estimated the mean and the variance of the received power at both a licensed pu and a malicious user using fenton approximation .
these estimations are compared with the probability ratio test on the received signals to detect pue attacks . however
, these techniques assume that the transmission power levels of attackers are significantly different than the fixed known transmission power levels of the pus . using neural networks , the authors in @xcite developed an approach to distinguish between pu and pue attacker .
the use of belief propagation was proposed in @xcite , in which each su calculates the belief based on the local and compatibility functions and the beliefs are exchanged between neighboring users and then used to detect pue attacks .
the work in @xcite adopts a game - theoretic approach in which the pue defense problem is a dog fight game between defending sus and the attacker .
previous supervised machine learning based techniques such as in @xcite require knowledge about users and authorized signals .
however , this kind of information is not always available in real cr implementations .
device fingerprinting @xcite is a process to extract device - specific information from the device transmitted signals .
these fingerprints form signatures that uniquely identify each individual pu or su in the system .
some of the features that can be used as fingerprints @xcite include the carrier frequency difference ( cfd ) , phase shift difference ( psd ) , and second - order cyclostationary feature ( socf ) .
since device fingerprinting requires no prior knowledge about pus and sus , unsupervised machine learning techniques can be used to detect pue attacks .
the authors in @xcite used an unsupervised bayesian mixture model to classify fingerprints into different groups .
each group represents a unique device in the system .
a pue attack is detected if two groups , physical devices , share the same i d , e.g. mac address .
another benefit of the unsupervised techniques is that they are passive and require no additional hardware to be added to the system unlike localization based techniques .
however , unsupervised fingerprinting techniques are applied at each time frame independently .
therefore , sufficient amount of fingerprints is required at each time frame , and it is required to have fingerprints for both the malicious attacker and the victim pu at the same time frame . the main contribution of this paper is to propose a novel transfer learning based framework in which knowledge from past time frames is transferred to current time frame to improve the final pue attacks detection decisions . to our knowledge
, this paper is the _ first to develop a transfer learning approach for device fingerprinting_. transferring knowledge is significant in fingerprinting scenarios in which an insufficient amount of fingerprints is available at the current time frame .
our proposed framework extracts abstract information about all su and pu devices in the cr system .
this abstract knowledge is used to boost the output of an unsupervised pue attacks detection algorithm .
the transfer learning part in the proposed framework updates the abstract knowledge after each time frame .
extensive simulations demonstrates that the transfer approach improves the efficiency of pue attacks detection especially in the case of insufficient availability of fingerprints .
results show that using the proposed framework the performance is enhanced with an average of @xmath0 for only @xmath1 of relevant information between the past knowledge and the current environment signals .
the rest of this paper is organized as follows .
section [ sec : model ] describes our system model . in section [ sec : method ] we explain in detail the proposed transfer learning framework .
simulation results and discussions are presented in section [ sec : results ] .
finally , conclusions are drawn in section [ sec : end ] .
consider a cr system composed of pus and sus , with the total number of pus and sus being @xmath2 .
each user has a unique identifier ( i d ) , such as a mac address .
secondary users ( sus ) can benefit from the spectrum only when all pus are idle .
hence , sus observe the signals of active pus to avoid using the spectrum when it is used by a licensed pu . in this model , we consider a pue attack scenario in which a malicious attacker copies the i d of one of the licensed pus , and , in turn , forces the sus to vacate the spectrum due to the fake pu - like signals sent from the attacker .
the pue attack causes a serious security threat and significantly degrades the total performance of the cr system . nevertheless , each device in the cr system , whether it is a pu or an su , transmits signals that include device - specific information , such information is the carrier frequency difference ( cfd ) , phase shift difference ( psd ) , and second - order cyclostationary feature ( socf ) .
such device - specific information is known to be unique to each transmitter @xcite , and hence can not be replicated by attackers .
the process to extract this device - specific information from the transmitted signals is called _ device fingerprinting _ @xcite . in device fingerprinting , the extracted device - specific information , or fingerprints , forms signatures that uniquely identify each device in the system .
therefore , even though the pue attacker shares the same i d with a real pu , the fingerprint data for the attacker is different from the fingerprint data for the attacked pu . although our model is applicable to any device fingerprint features , we restrict our attention to the cfd , psd , and socf . to detect pue attacks ,
a nonparametric bayesian model can be used , such as the one in @xcite . in this model
, fingerprinting data is gathered from all the devices in the system and grouped into different clusters . since fingerprints are unique for each device , each cluster of similar fingerprints belongs to one device only .
therefore , the number of clusters represents the real number of devices @xmath2 in the system .
the number of clusters is a key factor to detect the existence of a pue attack .
for example , if multiple clusters have the same i d , then a pue attack has occurred . to form the clusters
, one can adopt the nonparametric approach proposed by wood and black @xcite , that uses the infinite gaussian mixture model ( igmm ) . in this model
, the number of clusters , or mixtures , could be infinite .
using an igmm , one can cluster a given set of data points , such as the pus and sus fingerprints , by estimating a number of gaussian distributions , where each gaussian distribution represent a cluster .
the igmm therefore requires a mixture input and a set of hyper - parameters @xmath3 .
these hyper - parameters include priors on the mean and variance of the multivariate gaussian distributions .
to integrate the hyper - parameters out of the igmm , the mixture input needs to be generated from a distribution that is conjugate to the probability of the priors , @xmath4 with @xmath5 being the mean and @xmath6 being the variance .
hence , the dirichlet distribution is used for the mixing probability @xmath7 with @xmath8 being the mixing weights and @xmath9 being the mixing parameter .
the dirichlet distribution is given by @xmath10 for all @xmath11 where @xmath12 .
@xmath13 is the beta function , which can be expressed in terms of the gamma function as follows : @xmath14 the gibbs sampling process is then used to iteratively assign cluster labels to input data vectors , which in our case represent assigning fingerprints @xmath15 to devices .
gibbs sampler starts by initializing clusters for all the fingerprints @xmath15 .
subsequently , the sampler removes the cluster label for one of the fingerprints and calculates its probability for new cluster assignment .
the sampler iterates until convergence of all device assignments .
convergence is guaranteed only for the case of infinite number of fingerprints . in the real case in which a finite number of fingerprints is available ,
a maximum number of iterations for the gibbs sampler is chosen as a stopping criterion .
such a nonparametric bayesian method can be applied to a given set of fingerprints to detect pue attacks .
however , most such existing models , such as in @xcite , are static and they apply this method independently at every time period .
however , such static approaches require the cr system to have a sufficient amount of unique fingerprints for each device in the cr system to construct correct clusters for both the pus and the sus .
in such a nonparametric system , there might not be a sufficient amount of fingerprint data for each device in the system .
therefore , there is a need to handle the case of having an insufficient amount of fingerprints to efficiently detect pue attackers . to address this problem , next ,
we propose a transfer based framework in which fingerprinting results from past time frames are used to enhance the current pue attacks detection .
to enhance the fingerprinting process in a cr network , we use the powerful mathematical tool of transfer learning @xcite .
in particular , we propose a new transfer learning based framework for device fingerprinting , as illustrated in fig .
[ fig : method ] .
the proposed framework consists of two main components .
the first is the abstract knowledge database ( akd ) , which is able to gather general information about devices in the cr system from all time frames .
the second component is the transfer tool that updates the stored abstract knowledge database with the current time frame information .
the proposed framework receives the current fingerprint data from the environment , and suggests a clustering for the received fingerprint data based on the knowledge in the akd .
the clustering decisions from the proposed framework are merged with the clustering decisions of the static nonparametric bayesian model to form a final clustering decisions .
these final clustering decisions represent the assignment of fingerprint data to devices in the cr system .
the transfer tool updates the akd with the final clustering decisions for future pue attacks detection .
next , both the abstract knowledge database and the transfer learning algorithm will be described in detail .
the abstract knowledge database ( akd ) includes general information about the devices in the cr system starting from time @xmath16 , when the system was first deployed , and until the current time instant @xmath17 .
the information stored in the akd consists of a set of fingerprints @xmath18 , with _
db _ referring to the abstract knowledge database , grouped into multiple clusters @xmath19 , where each cluster represents a device in the cr system . the akd block in fig .
[ fig : method ] shows an example of 3 clusters of fingerprints .
hence , each fingerprint @xmath20 has a corresponding cluster label @xmath21 and a weight value @xmath22 .
weights represent the amount of confidence in assigning fingerprint @xmath23 to cluster @xmath24 . to reduce the number of fingerprints in the akd , a similarity metric
is introduced , which replaces two or more fingerprints @xmath25,@xmath26 that are nearly identical with new fingerprint @xmath27 , where @xmath27 minimizes the distances between @xmath25,@xmath28 : @xmath29 where @xmath30 is a distance measure and @xmath31 is the set of similar fingerprints : @xmath32 such that @xmath33 , @xmath34 .
hence , the new set of fingerprints @xmath35 in the akd equals to : @xmath36 .
for instance , if the distance measure is the l2-norm and we have two fingerprints @xmath37 with high similarity , @xmath38 then the new fingerprint @xmath39 using ( [ eq : simi ] ) will simply be the mean of both @xmath37 .
the distance threshold @xmath40 defines how close fingerprints need to be in order to be considered similar .
large values of @xmath40 allow more fingerprints to be merged and hence will result in a very generic abstract knowledge in the akd .
on the other hand , small values of @xmath40 tighten the similarity condition and result in a highly specific information in the akd .
when a new set of input fingerprint data @xmath41 arrives from the environment , the goal is to group these fingerprints into clusters . in order to group
the input fingerprints @xmath41 , the model compares these newly arriving fingerprints @xmath41 with the fingerprints @xmath18 stored in the akd to find the set of fingerprints @xmath42 which is as close as possible to the input fingerprints @xmath43 .
next , we use the clustering labels of @xmath23 as a suggested clustering labels for @xmath44 .
the model also uses the weights @xmath22 associated with each fingerprint @xmath23 in the akd to provide a confidence level along with each suggested clustering label .
a sigmoid logistic growth function is chosen to calculate the amount of confidence for each fingerprint in the database : @xmath45 where _
cf@xmath46 _ is the confidence weight for the fingerprint @xmath23 , @xmath22 is the weight associated with the fingerprint @xmath23 in the akd , @xmath47 and @xmath48 are the parameters of the logistic function that control how steep the logistic function is which in turn affects the level of confidence assigned to each weight .
the initial weights are chosen near the middle , based on the amount of confidence that we want to assign to the newly added fingerprints in the database .
set of fingerprint data @xmath41 from the environment set of labels @xmath24 and confidence values @xmath49 for all fingerprints @xmath50 calculate similarity between @xmath51 and @xmath52 @xmath53 ( of the @xmath52 similar to @xmath51 ) @xmath54 labels of @xmath55 @xmath56}$ ] clustering labels @xmath57 , confidence _ cf _ at specific time intervals , all the weights in the database are decreased by a fixed amount to reduce the confidence in old fingerprints which have not been used in the past several time frames .
the parameters @xmath47 and @xmath48 in ( [ eq : conf ] ) determine how long it takes for a fingerprint to be considered old and be removed from the akd .
therefore , the output of our proposed framework is a set of suggested clustering labels for the input fingerprints @xmath51 .
a confidence weight is also provided along with each suggested clustering label .
the suggested clustering labels from our proposed framework boost the clustering decisions of the bayesian model to construct final clustering labels .
the boosting process pseudocode is shown in algorithm [ alg : alg1 ] .
the purpose of the transfer tool is to use the final output decisions to update the akd . at each time frame , both the bayesian model and our proposed framework receive the current fingerprints @xmath15 from the environment , and group these fingerprints into clusters @xmath58 and @xmath24 respectively .
the _ decision _
block , shown in fig .
[ fig : method ] , merges the output of the bayesian algorithm @xmath59 and the output of the proposed framework @xmath60 into a final output decision @xmath61 , where @xmath15 is the input fingerprints , @xmath62 is the final cluster labels assignment .
since the final decisions have different labels than the labels in the akd , a mapping of labels is needed .
the transfer tool finds the right mapping of final decision labels @xmath62 to the labels in the akd @xmath24 by solving : @xmath63 where @xmath64 is the total number of fingerprints in @xmath41 , @xmath65 $ ] is the mapping vector between final cluster labels and cluster labels in the akd such that @xmath66 , where @xmath67 is the number of labels , for any two elements @xmath68 and @xmath69 of @xmath70 , @xmath71 is an indicator function in which @xmath72 when @xmath73 and @xmath74 when @xmath75 . for example , in the case of 2 clusters only , @xmath76 can take two possible values : @xmath77 $ ] or @xmath78 $ ] . after finding the right mapping of the labels from the final clustering output @xmath62 to the clustering labels
@xmath24 in the akd , the transfer tool updates the existing fingerprints @xmath79 in the akd with the final output labels of the input fingerprints @xmath51 . in the update process
, the input fingerprint data @xmath51 is divided into two categories : fingerprints with high similarity to the fingerprints @xmath23 in the akd , and fingerprints with low similarity to the fingerprints @xmath23 in the akd . in the high similarity case , we increase the weights @xmath22 for all the fingerprints @xmath23 in the akd that are of high similarity to the input fingerprints @xmath15 to give these fingerprints a higher confidence weight . as for the fingerprints with low similarity , the input fingerprints @xmath51 are considered new to the akd and are added to the akd with initial associated weights @xmath80 .
the algorithm used to update fingerprints in the akd is shown in algorithm [ alg : alg2 ] .
after specific time intervals , the weights for all the fingerprints in the database are decreased by @xmath81 to penalize for old unused fingerprints .
all the fingerprints @xmath23 with weights below a certain threshold are removed from the akd database .
fingerprint data @xmath51 from the environment , and similarity levels between fingerprints @xmath51 from the environment and fingerprints @xmath52 from the akd updated akd find the mapping of labels * _ m _ * ( between framework output labels and final output labels ) .
increase the weight of @xmath52 by @xmath82 decrease the weight of @xmath52 by @xmath83 add the new label mapping of @xmath51 to the akd with initial weight @xmath80 add the new fingerprint @xmath51 with its label to the akd with initial weight @xmath80 decrease all weights for all fingerprints in the akd by @xmath81
for our simulations , we consider the nonparametric bayesian method as our static pue attacks detection model , and we evaluate the benefits of using our proposed transfer based framework to enhance the detection of the pue attacks in the cr system . even though the hyper - parameters @xmath3 were integrated out of the iterative gibbs process as shown in section [ sec : model ] , an appropriate choice of hyper - parameters @xmath3 affects the speed of convergence of the gibbs sampler . the hyper - parameters @xmath3 consist of priors to the @xmath5 and @xmath84 of the multivariate gaussian distributions .
for example , the mean prior @xmath85 is usually set to be close to the mean value of all the fingerprints .
the hyper - parameter @xmath86 , which is proportional to prior mean for @xmath84 , determines how likely the fingerprints are to be variant from each others .
therefore , for features like signal amplitude , a large value of @xmath86 is chosen since signal amplitude feature has large variance between fingerprints .
on the other hand , frequency features tend to have small variance between fingerprints and hence a small value of @xmath86 is chosen for these features .
the dirichlet distribution is used in the generation of the fingerprints for the simulations .
random means and variances are used for each label .
.33 .33 .32 in this section , we evaluate the performance of the nonparametric bayesian model with and without our proposed transfer learning framework .
we use two main metrics for evaluation : ( i ) the hit rate of assigning the right label to each fingerprint , which is equals to the percentage of correctly assigned labels to fingerprints compared to the total number of fingerprints .
( ii ) missed attack rate , which is an indicator that missclassifying some fingerprints led into missclassifying a user in the system .
the second metric provides a better indicator on the performance enhancement due to using the transfer learning framework , this is due to the fact that even with high hit rate percentages , an attacker might not be detected if all the missclassified fingerprints belong to this attacker . on the other hand ,
even with low hit rate , accurate detection of attackers can be achieved if the missclassified fingerprints are distributed on various users .
the conducted experiments were repeated @xmath87 times and final results were averaged .
the number of users chosen is @xmath88 with total number of fingerprints ranging from @xmath89 to @xmath90 fingerprints .
an upper bound on the transfer learning approach is shown in the plots , which presents the maximum hit rate possible with perfect transferred knowledge .
this upper bound curve gives us an insight on the maximum performance that our proposed transfer approach can achieve . in fig .
[ fig : res1 ] , we show the hit rate resulting from the different learning approaches as the number of fingerprints varies . from this figure , we can see that our proposed framework achieved an average enhancement of @xmath0 in total hit rate ( for @xmath89 fingerprints ) with only @xmath1 of transferable knowledge .
[ fig : res1 ] also shows that the transfer learning upper bound curve yields to highest output hit rate if all the @xmath1 transferable knowledge were completely used .
the figure also shows that the maximum performance enhancement that can be reached through transfer learning is around @xmath91 making our model up to @xmath92 close to the optimal transfer case . as expected the more fingerprints we have that describes users the better the clusters are which yields to a higher total hit rate . fig .
[ fig : res2 ] shows the efficiency of using transfer learning as the number of fingerprints increases . as more fingerprints are available
, there is less need to transfer knowledge and , thus , the amount of performance improvement resulting from the proposed approach will naturally decrease .
this is due to the fact that having a large number of fingerprints will yield in a better clustering of labels by the detection algorithm which leaves small room of enhancement for the transfer approach . in fig .
[ fig : res3 ] , we show the number of times that the proposed transfer learning approach managed to correctly classify a pue attacker that was missclassified in the traditional detection approach .
as the number of fingerprints increases , missclassifying an attacker is less likely to happen .
this is due to the fact that even if some fingerprints were missclassified , having other correctly classified fingerprints insures that the attacker is correctly detected .
[ fig : res3 ] shows that for @xmath89 fingerprints the proposed transfer approach detects @xmath93 of the times an attacker which was not detected in the traditional non - transfer approach .
fig [ fig : res4 ] shows the effect of increasing the number of devices on the total hit rate for a fixed number of fingerprints for each user . in this experiment
, we obtain @xmath94 fingerprints for each user . as expected the hit rate decreases as the number of devices increases . for a fixed @xmath1
transferable information from past time frames , the figure shows that the effect of the proposed method increases as the number of devices increases , this is due to the fact that having 25 fingerprints for each user is sufficient in the 10 users case , and as the number of users increases , more fingerprints are needed to correctly cluster users ; and hence the more effective the transfer learning approach becomes .
in this paper , we have proposed a novel approach for performing device fingerprinting in wireless networks . in particular , we have introduced a novel framework , based on the machine learning tools of transfer learning , that enable static detection algorithms to benefit from past detection results .
we have applied the framework for fingerprinting pus and sus in a cognitive radio network .
our results have shown that the proposed framework enhanced the performance with an average of @xmath0 for only @xmath1 relevant information between the past knowledge and the current environment fingerprints . | primary user emulation ( pue ) attacks are an emerging threat to cognitive radio ( cr ) networks in which malicious users imitate the primary users ( pus ) signals to limit the access of secondary users ( sus ) .
ascertaining the identity of the devices is a key technical challenge that must be overcome to thwart the threat of pue attacks .
typically , detection of pue attacks is done by inspecting the signals coming from all the devices in the system , and then using these signals to form unique fingerprints for each device .
current detection and fingerprinting approaches require certain conditions to hold in order to effectively detect attackers .
such conditions include the need for a sufficient amount of fingerprint data for users or the existence of both the attacker and the victim pu within the same time frame .
these conditions are necessary because current methods lack the ability to learn the behavior of both sus and pus with time . in this paper ,
a novel transfer learning ( tl ) approach is proposed , in which abstract knowledge about pus and sus is transferred from past time frames to improve the detection process at future time frames .
the proposed approach extracts a high level representation for the environment at every time frame .
this high level information is accumulated to form an abstract knowledge database .
the cr system then utilizes this database to accurately detect pue attacks even if an insufficient amount of fingerprint data is available at the current time frame .
the dynamic structure of the proposed approach uses the final detection decisions to update the abstract knowledge database for future runs .
simulation results show that the proposed method can improve the performance with an average of @xmath0 for only @xmath1 relevant information between the past knowledge and the current environment signals . |
the importance of flux - line pinning in preserving the superconductivity in a magnetic field has been generally recognized since the discovery of type - ii superconductivity . but
till now the mechanism of flux - line pinning and creep in superconductors ( and particularly in the high-_@xmath0 _ superconductors ( htsc s ) ) is still a matter of controversy and great current interest , especially in the cases of strong competition between different types of pins .
one of the open issues in the field is the influence of _ isotropic _ point - like disorder on the vortex dynamics in the _ anisotropic _ washboard planar pinning potential ( ppp ) for the case of arbitrary orientation of the transport current with respect to the ppp `` channels '' where the _ guiding of vortices _ can be realized .
the importance of this issue may be substantiated by ubiquitous presence of point - like pins in those high- and low-_@xmath0 _ superconductors which were used so far for resistive measurements of the guided vortex motion@xmath1 .
the first attempt to discuss the influence of isotropic point - like disorder on the guiding of vortices was made by niessen and weijsenfeld@xmath2 still in 1969 .
they studied guided motion _ in the flux flow regime _ by measuring transverse voltages of cold - rolled sheets of a nb - ta alloy for different magnetic fields _ h _ , transport current densities _ j _ , temperatures _
t _ , and different angles @xmath3 between the rolling and current direction .
the ( _ h , j , t_,@xmath3)-dependences of the cotangent of the angle @xmath4 between the average vortex velocity @xmath5 and the vector * j * direction were presented . for the discussion , a simple theoretical model was suggested , based on the assumption that vortex pinning and guiding can be described in terms of an isotropic pinning force @xmath6 plus a pinning force @xmath7 with a fixed direction which was perpendicular to the rolling direction . the experimentally observed dependence of the transverse and longitudinal voltages on the magnetic field _ in the flux flow regime _ as a function of the angle @xmath3 was in agreement with this model .
unfortunately , in spite of the correct description of a geometry of the motive forces of a problem ( see below fig .
1 ) it was impossible within the flux flow approach@xmath2 to calculate theoretically the _ nonlinear _ ( _ j , t _ , @xmath3)-dependences of the average pinning forces @xmath8 and @xmath9 which determine the experimentally observed cot@xmath10 dependences .
the _ nonlinear guiding _ problem was exactly solved at first only for the washboard ppp ( i.e. for @xmath11 ) within the framework of the two - dimensional single - vortex stochastic model of anisotropic pinning based on the fokker - planck equation with a concrete form of the pinning potential@xmath12 .
two main reasons stimulated these theoretical studies .
first , in some htcs s twins can easily be formed during the crystal growth@xmath13 .
second , in layered htcs s the system of interlayers between parallel _ ab_-planes can be considered as a set of unidirectional planar defects which provoke the intrinsic pinning of vortices@xmath14 .
rather simple formulas were derived@xmath15 for the experimentally observable _ nonlinear _ even@xmath16 and odd@xmath17 ( with respect to the magnetic field reversal ) longitudinal and transverse magnetoresistivities @xmath18 as functions of the dimensionless transport current density @xmath19 dimensionless temperature @xmath20 and relative volume fraction @xmath21 occupied by the parallel twin planes directed at an angle @xmath3 with respect to the current direction .
the @xmath22-formulas were presented as linear combinations of the even and odd parts of the function @xmath23 which can be considered as the probability of overcoming the potential barrier of the twins@xmath15 ; this made it possible to give a simple physical treatment of the nonlinear regimes of vortex motion ( see below item ii.c ) .
besides the appearance of a relatively large even transverse @xmath24 resistivity , generated by the guiding of vortices along the channels of the washboard ppp , explicit expressions for _ two new nonlinear anisotropic hall resistivities _ @xmath25 _ and _ @xmath26 were derived and analyzed .
the physical origin of these _ odd _ contributions caused by the subtle interplay between even effect of vortex guiding and the odd hall effect .
both new resistivities were going to zero in the linear regimes of the vortex motion ( i.e. in the thermoactivated flux flow ( taff ) and the ohmic flux flow ( ff ) regimes ) and had a bump - like current or temperature dependence in the vicinity of highly nonlinear resistive transition from the taff to the ff .
as the new odd resistivities arose due to the hall effect , their characteristic scale was proportional to the small hall constant as for ordinary odd hall effect investigated earlier@xmath27 .
it was shown@xmath15 that appearance of these new odd @xmath28 contributions leads to the new specific angle - dependent `` scaling '' relations for the ppp which demonstrate the so - called anomalous hall behavior in the type - ii superconductors .
here we should to emphasize that the anomalous behavior of the hall effect in many high - temperature and in some conventional superconductors in the mixed state remains one of the challenging issues in the vortex dynamics@xmath29 .
the problem at issues includes several remarkable experimental facts : a ) the hall effect sign reversal in the vortex state with respect to the normal state at temperatures near @xmath30 and for moderate magnetic fields ; b ) the hall resistivity `` scaling '' relation @xmath31 exists with @xmath32 , where @xmath33 is the hall resistivity and @xmath34 is the longitudinal resistivity ; c ) the influence of pinning on the `` hall anomaly '' and scaling relation . assuming that the `` bare '' hall coefficient @xmath35 is constant , two different scaling laws have been derived earlier theoretically for different pinning potentials@xmath36 .
vinokur et al .
have shown@xmath37 that a scaling law @xmath38 ( where @xmath39 is the hall conductivity , @xmath40 , @xmath41 is the speed of light , @xmath42 is the magnetic field and @xmath43 is the magnetic flux quantum ) is the general feature of any isotropic vortex dynamics with an average pinning force directed along the average vortex velocity vector .
later it was shown@xmath15 that for purely anisotropic _
a_-pins that create a washboard planar pinning potential , the form of corresponding `` scaling '' relation is highly anisotropic due to the reason that pinning force for _
a_-pins is directed perpendicular to the pinning planes .
if @xmath3 is the angle between parallel pinning planes and direction of the current density vector @xmath44 , then for @xmath45 the scaling law has the form @xmath46 ( @xmath47 is the vortex viscosity ) which was interpreted previously@xmath15 as a scaling law with @xmath48 , whereas for @xmath49 the scaling relation is more complex@xmath15 .
the @xmath50 , as it is shown in this paper , can be presented as a sum of the three contributions with the different signs . the graphical analysis in sec .
iii of this paper represents a some range of the @xmath51-values where the theory predicts a nonlinear change of the @xmath50 sign .
let us consider another specific feature of the purely anisotropic guiding model@xmath12 . from the mathematical viewpoint , the nonlinear anisotropic problem , as solved in ref .
11 , reduces to the fokker - planck equation of the one - dimensional vortex dynamics@xmath52 because the vortex motion is unpinned in the direction which is parallel to the ppp channels . as a consequence , a critical current @xmath53 exists only for the direction which is strictly perpendicular to the ppp channels ( @xmath45 ) ; @xmath54 for any other direction ( @xmath55 )
however , the measurements of the magnetoresistivity show@xmath56 that @xmath57 for all @xmath3 ( although @xmath58 may be anisotropic ) .
so , in spite of some merits of a model with a washboard ppp , which was the first exactly solvable stochastic nonlinear model of anisotropic pinning , it can not describe the @xmath53-anisotropy of the experimentally measured samples .
due to this reason later it was suggested@xmath59 another simple model , which demonstrates this @xmath53-anisotropy for all @xmath3 on the basis of the bianisotropic pinning potential formed by the sum of two washboard ppp s in two mutually perpendicular directions .
in contradistinction to the nonlinear model with uniaxial ppp@xmath15 , this bianisotropic nonlinear model predicts a @xmath58-anisotropy and relates it to the guiding anisotropy , describing the appearance of two step - like and two bump - like singularities in the @xmath60 and @xmath61 ( hall ) resistive responses , respectively .
although several proposals to realize experimentally this bianisotropic model were discussed so far@xmath62 , the corresponding experiments , however , are still absent . at the same time , the experimental study of vortex dynamics in the ppp is always accompanied with a presence of a certain level of point - like disorder .
so , as far as the analysis of existing experimental data is concerned , none of the present theoretical studies in the limiting cases of purely anisotropic or isotropic pinning are sufficient .
the more general approach is needed .
the objective of this paper is to present results of a theory for the calculation of the nonlinear magnetoresistivity tensor at arbitrary value of competition between point - like and anisotropic planar disorder for the case of in - plane geometry of experiment .
this approach will give us the experimentally important theoretical model which demonstrates the @xmath53-anisotropy for all @xmath3 and predicts a nonlinear change of the @xmath26 sign at some set of parameters ( without change of the hall _ conductivity _ ) due a competition of the washboard ppp and a point - like disorder .
the organization of the article is as follows . in sec .
ii we derive main results of the @xmath63 pinning problem and consider two main limiting cases of purely @xmath64- or @xmath65-pinning . in sec .
iii we represent the graphical analysis of different types of nonlinear responses , in particular , the @xmath66 graphs of the @xmath67 magnetoresistivities and the resistive response in a rotating current scheme . in sec .
iv we conclude with a general discussion of our results .
the langevin equation for a vortex moving with velocity @xmath68 in a magnetic field @xmath69 ( @xmath70 , @xmath71 , @xmath72 is the unit vector in the @xmath73-direction and @xmath74 ) has the form @xmath75 where @xmath76 is the lorentz force ( @xmath77 is the magnetic flux quantum , @xmath41 is the speed of light ) , @xmath78 is the anisotropic pinning force ( @xmath79 is the washboard planar pinning potential ) , @xmath80 is the isotropic pinning force , induced by uncorrelated point - like disorder , @xmath81 is the thermal fluctuation force , @xmath82 is the vortex viscosity , and @xmath83 is the hall constant .
( with the unit vectors @xmath84 and @xmath85 ) associated with the ppp planes and the system of coordinates @xmath86 associated with the direction of the current density vector @xmath44 ; @xmath3 is the angle between the channels of the ppp and @xmath44 , @xmath4 is the angle between the average velocity vector of the vortices @xmath87 and the vector @xmath44 ; @xmath88 is the lorentz force ; @xmath89 and @xmath90 are the average isotropic and anisotropic pinning forces , respectively , @xmath91 is the average effective motive force for a vortex . here for simplicity we assume @xmath92 . ] for purely isotropic pinning ( i.e. for @xmath93 ) eq .
( [ f1 ] ) was earlier solved@xmath37 for @xmath94 , using the fact that @xmath95 where @xmath96 is velocity - dependent viscosity and @xmath97 . below we will show ( see eq .
( 8) and item d of sec .
ii ) , that the solution , obtained in ref .
17 , can be presented in terms of the probability function of overcoming the effective current- and temperature - dependent potential barrier of isotropic pinning @xmath98 , which is simply related to @xmath99 . in the absence of point - like disorder ( i.e. for @xmath100 ) eq .
( [ f1 ] ) was reduced to the fokker - planck equation , which was solved@xmath12 , assuming that the fluctuational force @xmath101 is represented by a gaussian white noise , whose stochastic properties are assigned by the relations @xmath102 where @xmath103 is the temperature in energy units . in what follows
we derive the solution of eq .
( [ f1 ] ) , using for @xmath80 the assumption ( [ f2 ] ) , which reduces eq .
( [ f1 ] ) to the equation @xmath104 where @xmath105 . using the result of ref .
11 , the selfconsistent solution of the eq .
( [ f4 ] ) can be represented as @xmath106 where @xmath107 is the probability of overcoming the ppp under the influence the effective moving force @xmath108 , @xmath109 and @xmath110 are the lorentz force components acting along the vector @xmath84 and @xmath85 , respectively , @xmath111 , @xmath112 and @xmath113 with an obvious condition @xmath114 .
( [ f5 ] ) can be rewritten as @xmath115 where @xmath116 and @xmath117 are corresponding right - hand parts of eqs .
( [ f5 ] ) . from eq .
( [ f6 ] ) we have @xmath118 where @xmath119 and we omitted for simplicity the symbol of averaging for @xmath68 .
then from eq .
( [ f7 ] ) follows that @xmath120 and thus it is possible to represent @xmath99 and @xmath121 in terms of @xmath122 : @xmath123\equiv\tilde{\eta}_{i}(f_{i})$ ] and @xmath124\equiv\nu_{i}(f_{i}).\ ] ] here @xmath98 has a physical meaning of the probability to overcome the effective potential barrier of isotropic pinning under the influence of effective ( @xmath125-dependent through the @xmath126-dependence ) force @xmath122 . then in terms of the @xmath98 eq .
( [ f6 ] ) takes the selfconsistent form @xmath127 which can be highly simplified for a small dimensionless hall constant @xmath128 . really , in this limit
@xmath129 , where @xmath130 , and the right - hand part of the eq . (
[ f6 ] ) becomes @xmath125-independent , i.e. is represented only in terms of the known quantities .
just in this limit all subsequent results of the paper will be discussed .
the average electric field induced by the moving vortex system is given by @xmath131 where @xmath84 and @xmath85 are the unit vectors in @xmath132- and @xmath133-direction , respectively . from formulas ( [ f9 ] ) and ( [ f10 ] ) we find the dimensionless magnetoresistivity tensor @xmath134 ( having components measured in units of the flux - flow resistivity @xmath135 ) for the nonlinear law @xmath136 @xmath137 the conductivity tensor @xmath138 ( the components of which are measured in units of @xmath139 ) , which is the inverse of the tensor @xmath134 , has the form @xmath140^{-1 } & n\epsilon \\ -n\epsilon & [ \nu_{i}(f_{i})\nu_{a}(f_{a})]^{-1 } \end { array}\right ) .
\end{array}\ ] ] from eqs .
( [ f11 ] ) and ( [ f12 ] ) we see that the off - diagonal components of the @xmath134 and @xmath138 tensors satisfy the onsager relation ( @xmath141 in the general nonlinear case and @xmath142 ) .
all the components of the @xmath134-tensor and the diagonal components of the @xmath138-tensor are functions of the current density @xmath143 through the external force value @xmath144 , the temperature @xmath103 , the angle @xmath3 , and the dimensionless hall parameter @xmath145 . for the following ( see item e.2 of sec .
ii ) it is important , however , to stress that the off - diagonal components of the @xmath138 ( i.e. the dimensional hall conductivity terms @xmath146 ) are not influenced by a presence of the @xmath65- and @xmath64-pins@xmath147 .
the experimentally measurable resistive responses refer to a coordinate system tied to the current ( see fig . 1 ) .
the longitudinal and transverse ( with respect to the current direction ) components of the electric field , @xmath148 and @xmath149 , are related to @xmath150 and @xmath151 by the simple expressions @xmath152 then according to eqs .
( [ f13 ] ) , the expressions for the experimentally observable longitudinal and transverse ( with respect to the @xmath44-direction ) magnetoresistivities @xmath153 and @xmath154 have the form : @xmath155 note , however , that the magnitudes of the @xmath156 , given by eqs .
( [ f14 ] ) , are , in general , depend on the direction of the external magnetic field @xmath157 along @xmath73 axis due to the @xmath158-dependence of the @xmath159 and @xmath160 forces in arguments of the @xmath161 and @xmath162 functions , respectively . in order to consider only @xmath163-independent magnitudes of the @xmath164- and @xmath165-resistivities we should introduce the even(+ ) and the odd(@xmath166 ) with respect to magnetic field reversal @xmath167 longitudinal and transverse dimensional magnetoresistivities , which in view of eqs .
( [ f14 ] ) have the form : @xmath168\nu_{i}(f_{i}),\\ \\ \rho_{\parallel}^{-}=\rho_{f}[[\sin^{2}\alpha+\nu_{a}(f_{lx})\cos^{2}\alpha]\nu_{i}^{-}(f_{i})+\\ \\ \qquad { } \nu_{i}(f_{i})\nu_{a}^{-}(f_{a})\cos^{2}\alpha].\\ \end{array}\ ] ] @xmath169\sin\alpha\cos\alpha,\\ \\ \rho_{\perp}^{-}=\rho_{f}[n\epsilon\nu_{a}(f_{lx})\nu_{i}^{2}(f_{i})+\\ \\ \qquad\{\nu_{a}^{-}(f_{a})\nu_{i}(f_{i})-\nu_{i}^{-}(f_{i})[1-\nu_{a}(f_{lx})]\}\times\\ \\ \qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad\sin\alpha\cos\alpha].\\ \end{array}\ ] ] here @xmath170 are the odd ( @xmath171 ) components of the functions @xmath98 and @xmath107 , and for @xmath172 we have the expansion in terms of @xmath173 : @xmath174.\ ] ] eqs . ( [ f15])-([f16 ] )
are accurate to the first order in @xmath173 and contain a lot of new physical information , which will be analyzed below ( see item e of sec .
ii ) . however , before this analysis it is instructive to discuss in short the main physically important features of two main limiting cases of purely anisotropic @xmath64-pinning and isotropic @xmath65-pinning , which follow from eqs .
( [ f15])-([f16 ] ) , when @xmath175 or @xmath176 , respectively .
setting @xmath175 we obtain rather simple formulas , which were derived firstly@xmath15 for the experimentally observable nonlinear even and odd longitudinal and transverse anisotropic magnetoresistivities @xmath177 as functions of the transport current density @xmath143 , dimensionless temperature @xmath178 and relative volume fraction @xmath179 , occupied by the parallel twin planes , directed at an angle @xmath3 with respect to the current direction : @xmath180,\quad \rho_{\perp a}^{+}=\rho_{f}(\nu_{a}^{+}-1)\sin\alpha\cos\alpha,\ ] ] @xmath181.\ ] ] here @xmath182 is considered as the probability of overcoming the potential barrier of the washboard ppp in the @xmath132-direction under the influence of the effective force @xmath183@xmath15 .
this @xmath184-function describes an essentially nonlinear transition from the linear low - temperature thermoactivated flux flow ( taff ) regime of vortex motion to the ohmic flux flow ( ff ) regime .
it is a step - like function of @xmath143 or @xmath178 for a small fixed temperature or current density respectively ( see figs . 4 , 5 in ref .
it follows from eqs .
( 18)-(19 ) that for @xmath185 the observed resistive response contains not only the ordinary longitudinal @xmath186 and transverse @xmath187 magnetoresistivities , but also two new components induced by the pinning anisotropy : an _ even transverse _ @xmath188 and an _ odd longitudinal _
component @xmath189 .
the physical origin of the @xmath190 ( which is independent of @xmath145 ) is related in an obvious way with the guided vortex motion along the " channels of the washboard pinning potential in the taff regime . on the other hand ,
the component @xmath189 is proportional to the odd component @xmath191 , which is zero at @xmath92 and has a maximum in the region of the nonlinear transition from the taff to the ff regime at @xmath192 ( see figs . 6 , 7 in ref .
11 ) the @xmath193-dependence of the odd transverse ( hall ) resistivity @xmath194 has contributions both , from the even @xmath195 and from the odd @xmath191 components of the @xmath196-function .
their relative magnitudes are determined by the angle @xmath3 and the effective hall constant @xmath145 .
note , that as the odd longitudinal @xmath197 and odd transverse @xmath198 magnetoresistivities arise by virtue of the hall effect , their characteristic scale is proportional to @xmath173 ( see eqs . ( 19 ) ) .
the appearance of these new odd hall contributions follows from emergence of a certain equivalence of @xmath199-directions for the case , where a guiding of the vortex along the channels of the washboard anisotropic pinning potential is realized@xmath200 at @xmath201 and leads to the new specific angle - dependent `` scaling '' relations for the hall conductivity@xmath15 for the case @xmath202 @xmath203 here the dimensionless hall constant @xmath173 is uniquely related to three experimentally observable nonlinear resistivities @xmath204 , and the `` scaling '' relation ( [ f20 ] ) depends on the angle @xmath3 .
this relation differs substantially from the power - law scaling relations , obtained in the isotropic case@xmath37 ( see below ) . in the particular case
@xmath45 we regain the results@xmath27 , specifically @xmath205 ( in ref .
10 @xmath206 ) , i.e. a linear relationship between @xmath207 and @xmath208 .
( 20 ) may be represented in another form @xmath209 which is more suitable for considering scaling relations in longitudinal ( @xmath49 ) and transverse ( @xmath45 ) lt - geometries of experiment@xmath15 . in these geometries
second term in the right hand side of eq .
( 21 ) is zero and we obtain that @xmath210 @xmath211 from eqs .
( 22)-(23 ) follows that @xmath212 may be interpreted as an _ effective _ hall conductivity in lt - geometries which is suppressed for @xmath49 ( @xmath213 ) and enhanced for @xmath45 ( @xmath214 ) in comparison with a bare hall conductivity @xmath215 .
the physical reason for this influence of @xmath184-function on the @xmath212 behavior in lt - geometries is simple .
namely , it appears as a result of the fact that in the case of anisotropic pinning the driving force @xmath183 , which determines the probability of overcoming the potential barrier ( and therewith also determines the magnitude of the component of the vortex velocity perpendicular to the channels of the ppp ) , is the sum of two forces .
the first of these is the transverse component of the lorentz force , @xmath109= @xmath216 , and the other is the _ transverse _ component of the hall force @xmath217 which is proportional to the longitudinal ( relative to the ppp planes ) component of the velocity of guided vortex motion .
this second force @xmath218 , which changes its sign ( relative to the sign of @xmath144 ) upon reversal of the sign of the external magnetic field , is the reason for appearance of new , hall - like in their origin , @xmath219-terms in the formulas for the resistive responses in eqs .
( 19 ) .
returning to the physics of suppression and enhancement of the @xmath212 in lt - geometries we should keep in mind that only longitudinal component of the vortex velocity ( with respect to the current direction ) @xmath220 is responsible for the appearance of the transverse hall voltage .
thus , in l - geometry @xmath220 and @xmath212 are suppressed by ppp - barriers , whereas in t - geometry @xmath220 is not influenced by them and @xmath212 looks like enhanced quantity . on the contrary , the behavior of the transverse component of the vortex velocity @xmath221 , which determines the longitudinal voltage , in lt - geometries is opposite . for this case
we put @xmath222 and from eqs .
( [ f15])-([f16 ] ) follows that @xmath223 where @xmath224 . from eqs .
( [ f21 ] ) the well - known scaling relation @xmath225 , derived firstly in ref . 17 , follows .
note that @xmath226 in this case , i.e. nonlinear resistive response is isotropic . equations ( [ f15])-([f16 ] ) for the magnetoresistivities @xmath227 at arbitrary value of competition between point - like and anisotropic planar disorder for the in - plane geometry of experiment can be represented in a more suitable form , if we take into account eqs .
( [ f18])-([f19 ] ) and ( [ f21 ] ) : @xmath228 @xmath229 @xmath230\}\sin2\alpha/2.\ ] ] here @xmath231 is the probability function @xmath161 of anisotropic argument @xmath232^{1/2}$ ] , the magnetoresistivity @xmath233 and the @xmath234-functions in eqs .
( [ f22])-([f24 ] ) are the same as those in item c of sec .
ii ; @xmath235 $ ] and @xmath236^{1/2}$ ] .
it is easy to check , that previous results of items c and d of sec .
ii follow from eqs .
( [ f22])-([f24 ] ) in the limits of purely anisotropic ( i.e. for @xmath175 , @xmath237 ) and isotropic ( i.e. for @xmath238 , @xmath239 ) pins . in this subsection
it must be suffice to discuss in short the main physically important features of these equations .
first of all , the magnetoresistivities @xmath227 can be found , if the @xmath162- and @xmath161- functions are known . moreover , the converse statement is also valid : it is possible to reconstruct these functions from ( @xmath143 , @xmath178 , @xmath42)-dependent resistive measurements , using only eqs .
( [ f22 ] ) , where the hall terms are ignored .
( [ f23 ] ) and ( [ f24 ] ) , which arise due to the hall effect , have a rather complicated structure , which reflects a more pronounced competition between isotropic and anisotropic disorder in the hall - mediated resistive responses .
let us outline the main new physical results , following from eqs .
( [ f22])-([f24 ] ) . for the discussion of the influence of point - like pins on the guiding of vortices in the anisotropic pinning potential it is sufficient to analyze eqs .
( [ f22 ] ) . whereas for the purely anisotropic pinning ( @xmath175 ) a critical current density @xmath240 exists only for direction , which is strictly perpendicular to the ppp ( @xmath45 ) and @xmath241 for any other direction ( @xmath242 ) due to the guiding of vortices along the channels of a washboard potential , in eqs .
( [ f22 ] ) the factor @xmath231 ensures that an anisotropic critical current density @xmath243 exists for arbitrary angles @xmath3 .
it is interesting , however , to note , that the angular dependence of the ratio @xmath244 , which determines the angle @xmath4 between @xmath44 and @xmath68 for @xmath64-pins in ref .
11 , according to the relation @xmath245 is not influenced by the isotropic disorder , because factor @xmath231 in eqs .
( [ f22 ] ) vanishes from eq .
( [ f25 ] ) .
physically it means , that character of anisotropy in the case of competition between @xmath65- and @xmath64-pinning is determined only by @xmath246\mathbf{f}_{lx}$ ] , ( see fig . 1 ) ,
i.e. by the average pinning force of the ppp .
isotropic pins influence only the magnitude of the average @xmath68-vector , because @xmath247 .
so , the polar resistivity diagram @xmath248 , which can be measured experimentally @xmath249 , is influenced by point - like pins , because from eqs .
( [ f11 ] ) follows , that @xmath250^{1/2}=\\
\\ \qquad { } \rho_{f}\nu_{i}(f_{i})(\sin^{2}\alpha+\nu_{a}^{2}\cos^{2}\alpha)^{1/2}.\\ \end{array}\ ] ] as it follows from eqs . ( [ f23])-([f24 ] ) ,
the odd longitudinal @xmath251 and transverse @xmath26 magnetoresistivities contain terms with the @xmath252-function .
they possess a highly anisotropic current- and temperature - dependent bump - like behavior .
they tend to zero in the linear regime of vortex motion . for @xmath253
these new terms disappear , because @xmath254 at these limits . as it was in the case of purely a - pinning ( see item c of sec .
ii ) , the appearance of these new odd hall contributions follows from the emergence of a certain equivalence of @xmath199-directions due to a guiding of vortices along the channels of the washboard pinning potential for the case with @xmath255 .
note also , that @xmath251 includes two terms with similar signs , whereas in @xmath26 there are terms with opposite signs .
the latter can give rise to the well - known sign change in the @xmath256-dependence of the hall resistivity below @xmath257@xmath14 .
( [ f22])-([f24 ] ) new anisotropic `` scaling '' relations for the dimensionless hall constant @xmath145 can be derived . for this purpose
we exclude @xmath258 from eqs .
( [ f22])-([f24 ] ) , for @xmath259 use eq .
( [ f17 ] ) ; and after some algebra in the limit @xmath260 we have : @xmath261\rho_{f}\nu_{i}^{2}}.\ ] ] it is easy to check that from eq .
( [ f27 ] ) follows scaling relations @xmath262 ( for @xmath65-pins at @xmath263 ) and eq .
( [ f20 ] ) ( for @xmath64-pins at @xmath264 ) as it follows from eqs .
( 25 ) and ( 27 ) , just the same `` scaling '' relations as given by eqs .
( 22 ) and ( 23 ) for _ a_-pins , exist also for ( _ _ i__+_a_)-pins ( with a replacement of corresponding _ a_-resistivities in eq .
( 8) by @xmath265 and @xmath266 ) .
physically it follows from the fact that point - like disorder does not change the angular dependence of the ratio @xmath267 , which determines the angle @xmath4 between @xmath44 and average velocity vector @xmath268 for @xmath64-pins@xmath15 , and influences only the magnitude of @xmath268@xmath147 .
in order to analyze different types of nonlinear anisotropic @xmath270-dependent magnetoresistivity responses , given by formulas ( [ f22])-([f24 ] ) , we should bear in mind that these responses , as is seen from formula ( [ f11 ] ) , are completely determined by the @xmath193-behavior of the functions @xmath107 and @xmath98 , having a sense of the probabilities to overcome the effective potential barriers of the @xmath64- and @xmath65-pins , respectively .
a simple analytical model for the calculation of the @xmath271-dependent @xmath269-functions was given earlier@xmath272 .
we will use for both @xmath161 and @xmath162 functions the one - dimensional periodic pinning potential @xmath79 ( see fig .
2 ) , which has a simple analytical form@xmath273 : @xmath274 where @xmath275 is the pinning force ( @xmath276 , where @xmath277 is the depth of the potential well and @xmath278 is the width of the well ) .
this form of @xmath79 allows to define as the properties of a given pinning center ( by the parameters @xmath279 and @xmath280 ) , as well as the density of such centers ( by the parameter @xmath281 , where @xmath282 is the period of the @xmath79 ) .
: @xmath282 is the period of the potential , @xmath278 is the width of the potential well , @xmath279 is the depth of the potential well , @xmath281 characterizes the concentration of the pinning planes . ]
calculation of the @xmath283 function on the basis of the pinning potential , given by eq .
( [ f28 ] ) , was done @xmath15 and can be represented here in the form@xmath15 @xmath284 where @xmath285 - 2f(f-1)\cosh[f(1-\varepsilon)/\theta\varepsilon-(1/\theta)]-\\ \\
\qquad\quad2f(f+1)\cosh[(f(1-\varepsilon)/\theta\varepsilon+1/\theta]\}/\sinh(f/\theta\varepsilon ) .
$ ] + here and below we have for the time being dropped the indices @xmath64 and @xmath65 from the physical quantities pertaining to pinning potentials @xmath286 and @xmath287 and formula ( [ f29 ] ) describes equally the pinning on both potentials . for convenience of qualitative analysis of the formulas following dimensionless parameters were used : @xmath288 is the effective motive force , which specifies its ratio to the pinning force @xmath276 , @xmath289 is the temperature .
the effect of the external force @xmath290 acting on the vortices consists in a lowering of the potential barrier for vortices localized at pinning centers and , hence , an increase in their probability of escape from them .
increasing the temperature also leads to an increase in the probability to escape of the vortices from the pinning centers through an increase in the energy of thermal fluctuations of the vortices .
thus the pinning potential of a pinning center , which for @xmath291 leads to localization of the vortices , can be suppressed by both an external force and by temperature .
a detailed quantitative and qualitative analysis of the behavior of @xmath292 as a function of all the parameters and its asymptotic behavior as a function of each are described@xmath15 . here
we will pay particular attention only to the typical curves of @xmath292 as a function of the parameters @xmath293 and @xmath178 , which describe the nonlinear dynamics of the vortex system as a function of the external force acting on the vortices in the direction perpendicular to the pinning centers and as a function of temperature ( see figs . @xmath294 and @xmath295 in ref .
11 ) . as we see from those figures , the form of the @xmath296 and @xmath297 curves
is determined by the values of the fixed parameters @xmath178 and @xmath293 .
the monotonically increasing function @xmath296 reflects the nonlinear transition of the vortex motion from the taff to the ff regime with the increasing external force at low temperatures ( @xmath298 ) , while at high temperatures ( @xmath299 the ff regime is realized in the entire range of variation of the external force ( even at small forces ) because of the effect of thermal fluctuation on the vortices .
the monotonically increasing function @xmath297 reflects the nonlinear transition from a dynamical state corresponding to the value of the external force at zero temperature to the ff saturation regime .
the width of the transition from the taff to the ff regime on the @xmath296 and @xmath297 curves depends on substantially different on the increasing of the parameters @xmath178 and @xmath293 , respectively .
namely , with increasing @xmath178 the function @xmath296 shifts leftward and becomes less steep ( see fig . 4 in ref .
that is , the higher the temperature , the smoother the transition from the taff to the ff regime and the lower the values of the external force , at which it occurs . with increasing
@xmath293 the @xmath297 curve also shifts leftward , it becomes steeper ( see fig . 5 in ref .
consequently , the greater the suppression of the potential barrier of the pinning center by the external force , the sharper the transition from the taff to the ff regime and the lower the temperature at which it occurs .
these graphs will be needed later on when we will discuss the physical interpretation of the observed guiding - depended resistive responses .
we also note that the dependence of the probability function @xmath300 on the concentration of pinning centers decreases monotonically from the value @xmath301 , which corresponds to the absence of pinning centers , and that it becomes steeper with decreasing fixed parameters @xmath293 and @xmath178 , owing to the growth of the probability density for finding the vortices at the pinning centers with decreasing temperature and external force .
let us turn to the dimensionless parameters by which one can in general case take into account the difference of the potentials @xmath302 and @xmath303 specifically , the difference of their periods @xmath304 , @xmath305 , the potential well depths @xmath306 , @xmath307 and the width @xmath308 , @xmath309 .
we introduce some new parameters : @xmath310 is the average concentration of pinning centers , @xmath311 is the average depth of potential well , @xmath312 , and @xmath313 , where the parameters @xmath314 and @xmath315 are measures of the corresponding anisotropies .
the temperature will be characterized by new parameters : @xmath316 and @xmath317 , which are the ratio of the energy of thermal fluctuations of the vortices to the average potential well depth @xmath306 and @xmath307 , respectively .
the current density will be measured in units of @xmath318 , where @xmath319 .
then the dimensionless parameters @xmath320 and @xmath321 , which specify the ratio of the external forces @xmath160 and @xmath322 to the pinning forces @xmath323 and @xmath324 ( @xmath162 and @xmath161 are the even functions of their arguments ) , we denote as @xmath325 and @xmath326 .
the values of the external force @xmath290 , at which the heights of the potential barriers @xmath306 and @xmath307 vanish at @xmath327 correspond ( at @xmath45 and @xmath49 ) to the critical current densities @xmath328 and @xmath329 respectively , where @xmath330 . in general case of nonzero temperature and @xmath331 it is possible to consider the angle - dependent crossover current densities @xmath332 and @xmath333 ( see below ) which correspond to change in the vortex dynamics from the taff regime to a nonlinear regime .
the condition , that determines the temperature region , in which the concept of critical current densities is physically meaningful is @xmath334 , because for @xmath335 the transition from the taff to the nonlinear regime is smeared , and the concept of critical current loses its physical meaning .
it is possible now to rewrite eqs .
( [ f15])-([f16 ] ) in the dimensionless form in order to represent them as functions of @xmath143 , @xmath178 , @xmath3 at given values of parameters @xmath336 , @xmath145 , @xmath337 , @xmath338 .
@xmath339,\ ] ] @xmath340\sin2\alpha/2,\ ] ] @xmath341+\nu_{a}^{-}(\tilde{f_{a}})\nu_{i}(f_{i})\cos^{2}\alpha,\ ] ] + @xmath342\}\sin2\alpha/2 , \end{array}\ ] ] + @xmath343 @xmath344 + and @xmath345 $ ] , @xmath346\sin2\alpha\}^{1/2}. \end{array}\ ] ] here + + @xmath347,@xmath348 , + + @xmath349 , @xmath350 , + @xmath351\pm\nu_{i}[\widetilde{f}_{i}(-n)]\}/2,&\\ \\ \quad & \nu^{\pm}_{a}(\widetilde{f}_{a})=\{\nu_{a}[\widetilde{f}_{a}(n)]\pm\nu_{a}[\widetilde{f}_{a}(-n)]\}/2 . &
\end{array}\ ] ] in eqs .
( 33)-(38 ) we also denoted @xmath352 and @xmath353 for simplicity . before following graphical analysis of the @xmath227 dependences given by eqs .
( [ f30])-([f33 ] ) , we should point out the magnitude of some parameters which will be used for presentation of the graphs . it is important to remind here that the parameter @xmath337 determines the value of anisotropy between @xmath161 and @xmath162 critical current densities , whereas the parameter @xmath338 describes the anisotropy magnitude of the width of nonlinear transition from the taff to the ff regime for @xmath161 and @xmath162 function .
more definitely , if @xmath354 , then @xmath355 and influence of the @xmath65-pins on the vortex dynamics decreases with @xmath337-increasing . for @xmath356 the situation is opposite and anisotropy effects may be fully suppressed with @xmath337-decreasing .
so , for the observation of pronounced competition between @xmath65- and @xmath64-pins @xmath357 should be taken .
the temperature dependences of the @xmath358 at small current densities under conditions of the presence both isotropic and anisotropic pinning potential were studied experimentally@xmath359 .
arrhenius analysis of these dependences within the frames of suggested here theoretical approach have shown that for the samples@xmath359 the @xmath360k , @xmath361k , @xmath362 nm , @xmath363 nm at @xmath364k . then for these samples @xmath365 , @xmath366 , @xmath367
it was also pointed out@xmath359 that the best fitting of the experimental and theoretical curves was established for @xmath368 , from which follows @xmath369 .
so for all graphs below we used @xmath370 , @xmath371 , @xmath372 , @xmath373 and if it is not pointed out specially , @xmath374 and @xmath375 . note also that for the even longitudinal resistivity @xmath376 and the even transverse resistivity @xmath377 for a small hall effect , terms proportional to @xmath378 are absent ( see eqs .
( [ f30])-([f31 ] ) ) and only contributions describing the competition between isotropic pinning and nonlinear guiding effect on the ppp in terms of the even @xmath161 and @xmath162 functions are presented . in order to discuss graphical @xmath66-behavior of the resistive responses we will use @xmath162 and @xmath161 functions of their arguments @xmath320 and @xmath321 , respectively , in the form given by eq .
( [ f29 ] ) . then these functions are , as a corresponding @xmath269-function@xmath15 , the step - functions in @xmath143 ( at fixed @xmath178 ) or in @xmath178 ( at fixed @xmath143 ) . for every of the @xmath269-functions
it is useful to determine the `` crossover current densities '' @xmath333 and @xmath332 , as those which correspond to the middle point of a sharp step - like nonlinear transition from the taff to the ff regime . as it follows from eqs .
( [ f34])-([f35 ] ) , we can present @xmath320 and @xmath321 as @xmath379 with @xmath380 , and @xmath381 with @xmath382 for @xmath383 and @xmath384 for @xmath385 ; @xmath386 for @xmath387 because eq .
( [ f35 ] ) can be presented in two equivalent forms , namely @xmath388 . .
in all following graphs the parameters @xmath370 , @xmath371 , @xmath372 , @xmath373 , @xmath374 , and @xmath375 ( unless otherwise stated ) . ] . ] . ] the behavior of @xmath389 function ( see fig .
3 ) is rather evident from the @xmath390 behavior .
namely , for all @xmath391 ( i.e. for @xmath392 ) ) the @xmath162 with a current increasing consistently follows next stages : a ) slow increasing at @xmath393 in the taff regime , where @xmath394 , b ) sharp step - like increasing with a width of the order of @xmath395 which corresponds to nonlinear transition from the taff to the ff regime , c ) second stage of slow increasing for @xmath396 which corresponds to the ff regime ( see also item c of sec .
it follows from the expression for @xmath397 that an increasing of @xmath3 and ( or ) @xmath337 leads to a broadening of the step of the order of @xmath395 and its shift to the larger current densities @xmath398 . the anisotropy of @xmath399 ( see eq .
( [ f35 ] ) and fig .
4 ) can be divided into two types : simple ( `` external '' ) which depends on @xmath400 , and more complex ( `` internal '' ) , given by @xmath401 .
the first ( external ) anisotropy stems from the `` tensorial '' @xmath3-dependence which exists also in the linear ( taff and ff ) regimes of the flux motion .
the second ( internal ) is through the @xmath3-dependence of @xmath162 , which in the region of transition from the taff to the ff regime is substantially nonlinear ( eq . ( [ f29 ] ) and fig . 3 ) .
the appearance of nonzero @xmath402 term in @xmath321 for @xmath403 physically describes the guiding of vortices along the channels of the ppp in the presence of @xmath65-pins for the current densities @xmath404 .
the influence of @xmath162-anisotropy on @xmath161 is different for different values of the angle @xmath3 ( see fig .
5 ) . for @xmath387 the anisotropy of @xmath401 does not influence the value of @xmath399 because @xmath405 in the expression for @xmath333 . on the contrary , for @xmath383
the influence of @xmath64-pins on @xmath406 is most effective for that range of current density , where @xmath407 , due to the inequality @xmath408 .
thus , the @xmath161 and @xmath162 as functions of the angle @xmath3 at @xmath409 behave themselves oppositely ( see figs . 3 , 5 ) : @xmath161 increases monotonically with @xmath3-increasing , whereas @xmath162 - monotonically decreases .
for @xmath410 and at small angles which meet the condition @xmath411 , the behavior of the @xmath161 and @xmath162 qualitatively similar in @xmath3 and opposite in @xmath337 . in case
where @xmath412 , the @xmath161 and @xmath162 behavior is qualitatively different and stems from the ( @xmath413)-dependences of the corresponding crossover current densities . in contradistinction to @xmath162 , the transition of @xmath161 from the taff to the ff depends weakly from @xmath3 and @xmath337 ; it moves to the lower current densities with @xmath337-increasing for @xmath387 and moves to the higher ones for @xmath383 .
in general , the @xmath162 behavior is more anisotropic than @xmath161 behavior .
the @xmath161 anisotropy appears only in the taff regime , whereas @xmath162 anisotropy exists as in the taff , as well in the ff regime . and
this anisotropy is greater in the current density as the angle @xmath3 is greater .
the @xmath161 and @xmath162 transition width at @xmath414 is defined by @xmath415 and @xmath416 parameters , respectively , and it increases for @xmath417 and @xmath418 .
now we are in a position to discuss the results of the presentation of eq .
( [ f22 ] ) in the form of graphs .
first we note that according to eqs .
( [ f22 ] ) , the even resistive responses can be represented as the products of corresponding isotropic and anisotropic @xmath269-functions .
for this reason the graphical analysis of the @xmath419 and @xmath420 , after the above - mentioned consideration of the @xmath421 ( see fig . 5 ) , can be reduced to the construction and analysis of the @xmath422 and @xmath423 graphs . for the value of the parameter @xmath374 . ] for the value of the parameter @xmath374 . ] for the value of the parameter @xmath424 . ]
let us begin with a discussion of @xmath425 behavior ( see eq .
( [ f18 ] ) and fig .
6 ) . for all @xmath426 due to the term @xmath402 in eq .
( [ f18 ] ) a critical current density @xmath240 exists only for direction , which is strictly perpendicular to the ppp ( @xmath45 ) ( as it was shown in item e.1 of sec .
ii ) and @xmath241 for any other direction ( @xmath242 ) due to the guiding of vortices along the channels of a washboard potential ( see also fig . 8 in ref . 11 ) . in the ff - regime the isotropization of the @xmath427 arises due to the vortex slipping over the ppp channels .
thus at small angles @xmath3 the @xmath162 function strongly influences the @xmath425 , whereas for @xmath428 this influence is not so effective due to the external anisotropy , which is proportional to the @xmath402 term .
returning now to the consideration of the @xmath419 graph we refer to the eq .
( [ f30 ] ) .
it is necessary to pay special attention to the taff behavior of these curves at small currents and temperatures , which follows from the full pinning of vortices by point - like pins .
this behavior is completely different ( for @xmath403 ) from the non - taff behavior of the corresponding graphs for the case of purely anisotropic pinning ( see fig .
8 in ref .
11 ) , which is provocated by the guiding of vortices along the channels of the ppp . at high current densities and ( or ) temperatures appears the ff regime , because the vortex motion transverse to the @xmath64-pins becomes substantial and longitudinal resistivity practically becomes isotropic . in these limiting cases the @xmath429 magnitudes are equal to unity ( fig .
7 ) . for the angles
@xmath331 the @xmath430 behavior follows substantially the properties of one multiplier . the qualitative behavior of these multipliers , depending on the @xmath143 and @xmath3 magnitude is very different as determined by different behavior of their crossover current densities @xmath431 and @xmath395 . the priority of a sharp rise of the appearance @xmath161 or @xmath162 functions depends on the competition between the crossover current densities @xmath431 and @xmath395 , respectively .
that is why it may appear a `` step '' on some of the @xmath430 curves ( for @xmath354 and @xmath432 ) when the next sequence of the vortex motion regimes is realized : a ) full @xmath65-pinning in the taff regime ( @xmath433 ) ; b ) nonlinear transition from the taff to the ff regime for @xmath65-pins ( @xmath434 ) , c ) practically linear the ff regime as a consequence of the guiding of vortices along the channels of the washboard ppp ( on the @xmath419 surface one can see the horizontal sections at @xmath398 , see figs .
7 , 8) ; d ) nonlinear transition to the ff regime of vortex motion transverse to the @xmath64-pins for @xmath435 and , at last , e ) a free ff motion for @xmath436 . with decreasing of the @xmath337 the a)-e )
corresponding regions along the current density axis @xmath143 can overlap each other and a common nonlinear transition appears instead of b)-d ) regions . for the limiting cases @xmath437 , a guiding of vortices is absent and the @xmath430 lt - behavior
is simply related to the @xmath161 and @xmath162 behavior .
if parameter @xmath416 is decreasing , then the width of the transition of @xmath162 from the taff to the ff is also decreasing .
such enhancement of the @xmath162 steepness leads to appearance of the minimum in @xmath3 for the @xmath419 graph ( see fig . 8) . ] .
pay attention to the inverted direction of the axes in comparison with fig.9 . ]
now we pass to a discussion of the @xmath423 and @xmath420 graphs . as it follows from eq .
( [ f18 ] ) , the @xmath438 and has a minimum in @xmath3 for all @xmath409 .
the @xmath439 reaches its maximal magnitude for @xmath440 due to the factor @xmath441 and realization of guiding in the taff regime for @xmath442 ( see .
9a in ref . 11 and fig.9 ) .
therefore , the most favorable angle for its observation is near @xmath443 . in considered case
the origin of this minimum has the same reason as a low ( @xmath143 , @xmath3)-behavior of the @xmath444 curves in fig .
7 , namely it stems from existence of the taff regime for the point - like pins at small @xmath143-values . as is seen in fig . 9
, the position and the magnitude of this @xmath445-minimum strongly depends on the @xmath3-value .
it is very much pronounced for @xmath354 and strongly suppressed for @xmath356 by influence of the @xmath65-pins . with increasing of the current density @xmath446 a position of the minimum in @xmath3 is shifting due to the competition of two multipliers in the @xmath423 expression eq .
( [ f18 ] ) ) : @xmath447 is decreasing for @xmath448 , whereas @xmath449 is increasing with @xmath3-increasing for @xmath409 , and decreasing for @xmath143-increasing for @xmath414 due to the transition to the ff regime . for all
@xmath3 and current densities @xmath396 the @xmath450 , and for this reason @xmath451 .
the @xmath337-influence is defined by @xmath332 and determines the region of appearance of a small value of the @xmath439 for the current densities @xmath398 . since the @xmath452 , according to eq .
( [ f31 ] ) , is the product of the @xmath439 and @xmath453 , so this graph ( see fig .
10 ) can be reduced to the product of the graphs in fig . 5 and fig .
the transition from the taff to the ff regime is highly anisotropic in @xmath3 ; this causes a shift of the maximal @xmath420 magnitude in the direction of a small angle @xmath383 for the @xmath409 . that is why in view of @xmath65-pinning presence the @xmath420 , as distinct from @xmath454 , has the minimum both in @xmath3 and in @xmath143 .
this statement follows from the fact that influence of @xmath65-pinning leads to the @xmath455 for @xmath456 due to the @xmath457 . for the current densities
@xmath410 the @xmath420 behavior is determined exclusively by the above - mentioned @xmath423 behavior . before following discussion of the odd resistive responses we should remind the reader about the bump - like behavior of the current and temperature dependence of the @xmath170 functions ( see figs . 6 and 7 in ref .
11 ) , because @xmath458 and @xmath459 functions , as it follows from eqs .
( [ f32])-([f33 ] ) , give an important contribution to the odd responses .
the @xmath460 and @xmath461 curves for the case of @xmath462 in fact are proportional to the derivatives of the corresponding @xmath463 and @xmath464 curves , which have a step - like behavior as a function of their arguments ( see ref . 11 for the detailed discussion of this point and eq .
( [ f17 ] ) in this paper ) . as the @xmath465 and @xmath466 resistivities given by eqs .
( [ f23])-([f24 ] ) arise by virtue of the hall effect , their characteristic scale is proportional to @xmath462 , as for eqs .
( [ f19 ] ) for purely anisotropic pins . .
] . ]
the position of the characteristic peak in the @xmath458 and @xmath459 functions is different for @xmath467 , because parameter @xmath337 determines the anisotropy of the critical current densities for @xmath65- and @xmath64- pins .
so , if @xmath337 is not very close to the unity , the position of the @xmath65- and @xmath64- peaks can not coincide , and in this case the current and temperature odd resistive dependences @xmath468 can have a bimodal behavior . for the @xmath465 curves such dependences will correspond to existence of the resistive `` steps '' on the @xmath376 curves
( see fig .
7 ) , because for @xmath462 we can consider the @xmath465 dependences as derivatives of the @xmath376 curves . from this viewpoint
it is easy to understand the previous assertion in item e.2 of sec .
ii that @xmath465 includes two terms ( every proportional to the @xmath458 and @xmath459 , respectively ) with similar signs .
now we will discuss the @xmath259 and @xmath258 as a function of ( @xmath469 ) and the parameter @xmath337 in detail .
really , due to the smallness of the hall constant , the @xmath259 and @xmath258 tend to zero in the regions of the linear taff and ff regimes of the @xmath162 and @xmath161 function , respectively .
the @xmath259 and @xmath258 functions have a sharp peak ( see fig . 11 , 12 ) in the region of sharp change of the @xmath162 and @xmath161 increasing ( for @xmath398 or @xmath470 , respectively ) . with @xmath3- and @xmath337-increasing the width and
the height of the @xmath259 maximum also increases with simultaneous shift of the maximum to the higher current densities due to the relation @xmath471 .
the @xmath258 peak is located in the angle range @xmath472 , which corresponds to a change of the angular dependence of the crossover current density @xmath333 from the angles @xmath473 to the angles @xmath383 ( see .
the @xmath258 maximum shifts to a smaller current densities with @xmath337-increasing due to the @xmath474 , .
the magnitudes of the @xmath259 and @xmath258 are compete by an order of magnitude for @xmath472 and all @xmath337-values which satisfy a condition @xmath475 . .
the characteristic minimum ( which is shown by the arrow ) is in the region @xmath476 and @xmath477 . the minimum is shown as two neighboring minimums due to the step - like behavior of the calculation .
pay attention to the inverted direction of the axes in comparison with fig.13 . ]
now let us discuss a graphical presentation of the eq .
( [ f32 ] ) , which can be represented as @xmath478 , where @xmath479 , and @xmath480 . taking into account
that every factor in the @xmath481 and @xmath482 is positive ( see figs . 5 , 6 , 11 , 12 ) , we can conclude that @xmath483 for all values of the @xmath484 . proceeding to the analysis of the @xmath481 and @xmath482 @xmath485-behavior in details we consider first those limiting cases in which @xmath64- or @xmath65- pinning is dominant i.e. @xmath486 or @xmath450 , respectively .
if @xmath64-pinning is dominant ( i.e. for @xmath487 ) , then @xmath488 , and eq.([f32 ] ) has the form @xmath489 . for the opposite case ( i.e. for @xmath356 ) ,
conversely , @xmath490 , and @xmath491 .
the @xmath492 graph presentation is especially simple because it may be depicted with the aid of figs . 5 , 11 , 12 . in the general case ,
i.e. for @xmath357 , we should consider the @xmath481 and @xmath482 separately because dominant type of pinning is absent .
the @xmath481 is proportional both @xmath258 , which is nonzero for @xmath476 and @xmath493 ( see fig .
12 ) , and the factor @xmath208 with a graph , shown in fig . 6 . as a result
, the @xmath481 has a sharp maximum for the @xmath476 and @xmath493 .
the second term @xmath482 is proportional both the factor @xmath259 and the factor @xmath494 .
the contribution of the first factor is maximal for @xmath495 and current densities @xmath496 , whereas the @xmath494 contribution is maximal for @xmath472 and @xmath493 .
therefore these factors compete so that the resulting maximum of @xmath482 shifts from @xmath495 to the @xmath497 .
it is relevant to note that the condition @xmath498 for @xmath357 and @xmath497 is always fulfilled .
that is why the maximal contribution of the @xmath482 is realized for @xmath499 because in this region of the current densities the @xmath500 for @xmath497 .
therefore the @xmath501 behavior is determined mainly by the @xmath482 behavior , and the @xmath481 contribution is essential for @xmath472 and @xmath477 .
the @xmath502 dependence is the most complicated . for the sake of simplicity
the analysis we represent the @xmath503 as a sum @xmath504 $ ] , where @xmath505 , @xmath506 , @xmath507 .
first we consider the limiting cases of purely isotropic or anisotropic pinning ( @xmath508 or @xmath500 , respectively )
. for @xmath65-pinning we have @xmath509 , from which follows ( vinokur et al.@xmath37 ) a scaling relation @xmath510 . for the case of purely anisotropic pinning @xmath511 , and
the scaling relation is @xmath512 ( see also ref .
now we consider every term in the @xmath502 in detail .
the @xmath513 contribution can be reduced in fact to the multiplication of the graph in fig . 3 by the graph in fig .
5 squared ; the result is essentially nonzero for @xmath514 .
the @xmath515 contribution was described above ( see the @xmath482 term in the @xmath501 without taking into account the @xmath400 anisotropy ) .
note also that both terms ( @xmath513 and @xmath515 ) are positive for @xmath516 .
the @xmath517 behavior is of great interest because the @xmath518 for @xmath516 .
let us consider the cases @xmath354 and @xmath356 , which correspond to the @xmath64- , or @xmath65-pinning domination , respectively .
then , for @xmath519 : \a ) for @xmath356 we have @xmath520 and the sharp maximum of the @xmath458 is suppressed by the factor @xmath521 . as a result , the @xmath517 contribution can be ignored .
\b ) for @xmath522 the opposite inequality follows , i.e. @xmath498 . then for @xmath523 the @xmath517 term is dominant because @xmath394 and @xmath524 in this @xmath66-region ( see fig . 3 and fig . 5 ) . as a result ,
the @xmath525 change the sign for @xmath523 and @xmath472 . since the scale of the @xmath526 , the amplitude of the minimum is small in comparison with the @xmath503 magnitude .
thus , a competition of the @xmath64- and @xmath65-pinning leads to the qualitatively important conclusion that the @xmath503 can change its sign at a certain range of @xmath527-values , namely for @xmath528 , @xmath472 , and @xmath354 .
an experimental study of the vortex dynamics in @xmath529 crystals with unidirectional twin planes was recently done using a modified rotating current scheme@xmath530 . in that scheme it was possible to pass current in an arbitrary direction in the @xmath531 plane of the sample by means of four pairs of contacts placed in the plane of the sample .
two pairs of contacts were placed as in the conventional four - contact scheme , and the other two pairs were rotated by @xmath532 with respect to the first ( see the illustration in fig .
@xmath533 of ref .
4 ) . by using two current sources connected to outer pair of contacts , one can continuously vary the direction of the current transport in the sample . by simultaneously measuring the voltage in the two directions ,
one can determine directly the direction and magnitude of the average velocity vector of the vortices in the sample as a function of the direction and magnitude of the transport current density vector .
this made it possible to obtain the angular dependence of the resistive response on the direction of the current with respect to the pinning planes on the same sample .
the experimental data@xmath530 attest to the anisotropy of the vortex dynamics in a certain temperature interval which depends on the value of the magnetic field .
a rotating current scheme was used@xmath534 to measure the polar diagrams of the total magnetoresistivity @xmath248 , where @xmath535 is the absolute value of the magnetoresistivity , @xmath536 and @xmath537 are the @xmath132 and @xmath133 components of the magnetoresistivity in an @xmath199 coordinate system , and @xmath3 is the angle between the current direction and the @xmath538 axis ( parallel to the channels of the @xmath64-pinning centers ) . in the case of a linear anisotropic response
the polar diagram of the resistivity is an ellipse , as can easily be explained . in the case of a nonlinear resistive response
the polar diagram of the resistivity is no longer an ellipse and has no simple interpretation . in this subsection
we carry out a theoretical analysis of the polar diagrams of the magnetoresistivity @xmath539 in the general nonlinear case in the framework of a stochastic model of @xmath540 pinning .
this type of angular dependence @xmath248 is informative and convenient for theoretical analysis . for a sample with specific internal characteristics of the pinning ( such as @xmath337 , @xmath416 , @xmath415 , and @xmath314 ) at a given temperature and current density the function @xmath248 is contained by the resistive response of the system in entire region of angles @xmath3 and makes it possible to compare the resistive response for any direction of the current with respect to the direction of the planar pinning centers .
in addition , in view of the symmetric character of the @xmath248 curves , their measurements makes it possible to establish the spatial orientation of the system of the planar pinning centers with respect to the boundaries of the sample if this information is not known beforehand . for a sequence of the parameter @xmath143 : 0.63 ( 1 ) , 0.65 ( 2 ) , 0.75 ( 3 ) , 1.00 ( 4 ) , 1.50 ( 5 ) , 1.92 ( 6 ) , 2.00 ( 7 ) , 2.50 ( 8) , 4.00 ( 9 ) , 20.0 ( 10 ) for @xmath374 . ] for a sequence of the parameter @xmath143 : 0.63 ( 1 ) , 0.65 ( 2 ) , 0.75 ( 3 ) , 1.00 ( 4 ) , 1.50 ( 5 ) , 1.92 ( 6 ) , 2.00 ( 7 ) , 2.50 ( 8) , 4.00 ( 9 ) , 20.0 ( 10 ) for @xmath541 . ] now for analysis of the @xmath248 curves we imagine that vector * j * is rotated continuously from an angle @xmath49 to @xmath45 .
the characteristic form of the @xmath248 curves will obviously be determined by the sequence of dynamical regimes through which the vortex system passes as the current density vector is rotated . by virtue of the symmetry of the problem
, the @xmath248 curves can be obtained in all regions of angles @xmath3 from the parts in the first quadrant .
we recall that in respect to the two systems of pinning centers it is possible to have the linear taff and ff regimes of vortex dynamics and regimes of nonlinear transition between them .
the regions of nonlinear transitions are determined by the corresponding values of the crossover current densities @xmath542 and @xmath397 .
now let us consider the typical @xmath248 dependences which are presented in fig .
15 and 16 for a sequence of a current density magnitude .
we remind that the polar diagram graphs represented below are constructed , as the previous graphs in figs .
3 - 7 , 9 - 14 for the next values of parameters : @xmath370 ( i.e. for the case with dominant @xmath64-pins ) , @xmath371 , @xmath372 , @xmath373 , @xmath375 , @xmath374 ( fig .
15 ) , and @xmath541 ( fig .
note that @xmath248 is the product of two multipliers : one is the @xmath453 dependence , which was earlier studied in fig .
4 of item c.1 of sec .
iii , and other is the @xmath543 factor , which qualitative behavior is close to the @xmath422 dependence ( see fig . 6 in item c.2 of sec .
let us analyze the @xmath248 behavior for the series of values of the current density @xmath143 .
when the angle @xmath3 changes from @xmath544 to @xmath545 the function @xmath248 grows monotonically from @xmath546 to @xmath547 . in fig .
15 curves 1 - 6 of the function @xmath248 have the shape of the 8-figure drawn along @xmath548-axis ( strongly elongated for the curves 1 , 2 ) .
this anisotropy can be determined by the relation of the magnitudes of the half - axis at the direction @xmath49 to the transverse half - axis for any fixed magnitude of the current density .
the curves 1 - 6 of the @xmath549 graph has the 8-form elongated along @xmath548-axis .
it is caused by the step - like behavior of the @xmath161-function , corresponding for the curves 1 , 2 to the crossover from the taff to the ff regime .
that is why the magnitude of the @xmath550 for the curve 2 is rather greater than for the first one . with @xmath3-increasing
the @xmath161-function is in the taff - region ( see fig.5 ) , which provocates the @xmath551 in the case where the condition @xmath552 is satisfied .
therefore , with @xmath143-increasing the magnitude of the angle @xmath3 , which separates the taff and the ff regions of the @xmath161-function at a fixed value of the current density , decreases to the @xmath553 . as the @xmath554 is in the ff region ( i.e. @xmath555 ) , so the anisotropy of the 8-curve decreases for curves 3 - 6 .
the @xmath549 behavior of the curves 5 - 6 is more isotropic in the region @xmath383 than behavior of the curves 1 - 4 . if the condition @xmath556 is satisfied , an appearance of the nonzero resistance in corresponding region follows .
its magnitude is smaller than @xmath550 for the curves 7 , 8 , 9 and practically is equal to the @xmath550 for the curve 10 .
note , that for the @xmath557 and @xmath558 one can see the minimum , which decreases with @xmath143-increasing and disappears in the case where the condition @xmath559 is satisfied .
so , for large magnitudes of the current densities the @xmath248 behavior becomes more isotropic .
it is necessary to pay attention for the @xmath549 behavior in the case where @xmath560 ( see fig .
16 ) for the same series of the magnitudes of the current densities .
the behavior of the curves 6 , 7 , 8 , 9 differs from the above - mentioned case , but the behavior of the curves 1 - 5 , 10 retains the same . this fact is caused by the influence of the parameter @xmath416 on the @xmath161 behavior only in the area of its sharp step - like behavior at the @xmath561 .
note , that the @xmath161 contribution is dominant in the region @xmath553 as well as the above - mentioned anisotropy of the @xmath423 ( see fig . 5 and fig . 7 ) .
as decreasing of the @xmath416 causes the more narrow crossover from the taff- to the ff - regime , the @xmath406 has a minimum at fixed magnitude of the current density .
the magnitude of this minimum decreases with the @xmath143-increasing and the minimum shifts from the @xmath562 to the @xmath563 .
the influence of the parameter @xmath337 acts on the crossover current densities @xmath431 and @xmath395 only quantitatively , but does not change an evolution of the curves 1 - 10 qualitatively . let us examine theoretically in our model a new type of the experimental dependence , recently studied@xmath534 for @xmath564 , where @xmath565 is the angle between @xmath44-vector and the electric field vector @xmath566 measured at fixed values of the current density and temperature .
taking into account that in the _ xy _ coordinate system the magnetoresistivity components are @xmath567 , @xmath568 , we obtain the following simple relation : @xmath569 , or @xmath570 note , that the @xmath161 term , describing the @xmath65-pinning , is absent in eq.([f36 ] ) .
then it follows from the latter that the @xmath571 , i.e. the @xmath572 function can be found from the experimental dependence @xmath564 .
unfortunately , the dependence @xmath564 for the series of the temperature values was experimentally found@xmath573 so far only for the ff - regime ( see .
fig . 2 in ref . 4 ) .
the @xmath574 dependence is presented in fig .
it shows all changes in the @xmath574 behavior also for the taff - regime . for a sequence of the parameter @xmath143 : 1 ( 1 ) , 1.7 ( 2 ) , 2.2 ( 3 ) , 3.5 ( 4 ) , 20 ( 5 ) for @xmath575 . ]
let us analyze the eq .
( [ f36 ] ) in detail .
the @xmath574 is the odd function of the angle @xmath3 , and its magnitude increases monotonically with the @xmath3-increasing for all values of the @xmath143 due to the monotonical decreasing of the @xmath389 function ( see .
it follows from eq .
( [ f36 ] ) that the period of the function @xmath564 is equal to @xmath576 .
one more important limiting case is realized for @xmath450,which corresponds to the limit of isotropic pinning .
depending on the inequality between the @xmath143 magnitude and the crossover current density @xmath577 , one can separate two regions where the @xmath574 behavior is qualitatively different . if @xmath578 is the argument of the arctangent function in eq .
( [ f36 ] ) , then in that region @xmath484 , where the inequality @xmath410 is true ( the ff regime for @xmath389 , see also fig .
3 ) , the magnitude of the @xmath579 as @xmath580 . and for the case @xmath442 ( the taff regime of the @xmath389 ) the value @xmath581 , as @xmath582 .
note , that the parameter @xmath416 influences the @xmath574 by changing the character of the step - like crossover of the @xmath389 ( the smaller the @xmath416 , the sharper the crossover ) .
the value of the parameter @xmath337 , as well as above - mentioned , determines the magnitude of the @xmath332 ( and , therefore the position of the boundaries in @xmath143 of the regions of quite different @xmath564 behavior ) at fixed @xmath3 . under the critical current density we mean the current density , which corresponds to the electric field strength on the sample @xmath583 .
let us determine the @xmath584 behavior graphically by crossing the @xmath585 graph and the plain @xmath586 in the polar coordinates .
for all angles @xmath3 the point of crossing for these graphs determines the critical current density magnitude for the defined direction , and the crossing line of the graphs presents the dependence @xmath584 .
let us remind the reader that as in above - mentioned sections , in the nonlinear law @xmath585 we measure @xmath143 and @xmath539 in the values of the @xmath587 and @xmath588 , respectively .
that is why the @xmath589 magnitude we have to measure in the @xmath590 . as well as in item b of sec .
iii we use the data from ref . 8 , where for the niobium samples @xmath591 ohm@xmath592 cm , @xmath593 gs , @xmath594 17 kgs , @xmath595ohm@xmath592 cm , @xmath596 , and @xmath597 . therefore , @xmath598 , and for @xmath599 we have to cross the dimensionless @xmath600 graph by the plain @xmath601 . for the parameter pairs
: @xmath602 , @xmath337=1.6 ( 1 ) ; @xmath602 , @xmath337=3 ( 2 ) ; @xmath603 , @xmath337=1.6 ( 3 ) . ]
now we will discuss the @xmath584 as a function of @xmath604 , and @xmath336 in detail .
the @xmath584-anisotropy can be determined by the relation of the magnitudes of the half - axis at the direction @xmath45 to the transverse half - axis for any fixed magnitude of the parameters @xmath337 , @xmath605 .
the @xmath584 decreases monotonically from @xmath606 with @xmath3-increasing and has a minimum for @xmath49 .
it is caused by the fact that , as it was shown in item c.1 of sec .
ii , the @xmath64-pinning ( with high values of the @xmath395 for @xmath354 ) does not influence the @xmath65-pinning for @xmath49 .
therefore , the inequality for the crossover current densities @xmath498 for @xmath354 leads to the corresponding inequality for the critical current densities @xmath607 .
the @xmath337 influences the @xmath584 behavior ( as in item d.1 of sec .
ii ) only quantitatively : with @xmath337-increasing the ratio @xmath608 grows and visa versa .
it is caused by the @xmath606-increasing and @xmath609-decreasing due to the @xmath3-behavior of the corresponding crossover current densities @xmath333 and @xmath332 .
the smaller the @xmath610 , the sharper the crossover between the @xmath584 regions of slowly and quickly decreasing as a function of the @xmath3 . with @xmath605-increasing the nonlinear law
@xmath611 is satisfied for the larger values of the current density .
that is why with @xmath3-increasing from @xmath544 to @xmath612 values ( for which the condition @xmath613 is satisfied ) the @xmath162-function is in the ff regime and @xmath584 decreases slowly .
when the condition @xmath614 is true the @xmath162-function has a step - like crossover from the ff to the taff regime and @xmath584 decreases quickly .
so , the @xmath612 behavior as a function of the parameters @xmath337 and @xmath605 is qualitatively different : it increases with @xmath605-increasing and decreases with @xmath337-increasing . on the increase of the @xmath605 by the several orders of magnitude the @xmath584 curve degenerates into a circumference due to the isotropization of the @xmath333 and @xmath332 behavior for the high @xmath143-values
otherwise , with @xmath605-decreasing the @xmath584 curve degenerates into a narrow loop , because the @xmath333 and @xmath332 behavior for a small @xmath143 is very anisotropic .
in the present work we have theoretically examined the strongly nonlinear anisotropic two - dimensional single - vortex dynamics of a superconductor with coexistence of the anisotropic washboard ppp and isotropic pinning potential as function of the transport current density @xmath143 and the angle @xmath3 between the direction of the current and ppp planes at a fixed temperature @xmath178 . the experimental realization of the model studied here can be based on both naturally occurring@xmath615 and artificially created@xmath616 systems with @xmath63 pinning structures .
the proposed model has made it possible for the first time ( as far as we know ) to give a consistent description of the nonlinear anisotropic current- and temperature - induced depinning of vortices for an arbitrary direction relative to the anisotropy of the washboard ppp . in the framework of this model
one can successfully analyze theoretically certain observed resistive responses which are used for studying anisotropic pinning in a number of new experimental techniques@xmath573 ( the polar diagram of @xmath248 , the @xmath564 curve described by formula eq .
( [ f36 ] ) ) as well as new hall responses specific for the @xmath63 pinning problem .
a quantitative description of the anisotropic nonlinear resistive properties of the problem under study is done in the framework of the stochastic model on the basis of the fokker - planck approach .
the main nonlinear components of the problem are the anisotropic @xmath107 and isotropic @xmath231 probability functions for the vortices to overcome the potential barriers of @xmath64- and @xmath65-pinning centers under the action of anisotropic motive forces @xmath160 and @xmath322 , respectively .
the latter include both the `` external '' parameters @xmath617 and the `` internal '' parameters @xmath618 which describe the intensity and anisotropy of the pinning . as can be seen from eqs .
( [ f30])-([f33 ] ) , the magnetoresistivities @xmath619 are , in general , nonlinear combinations of the experimentally measured @xmath161 and @xmath162 functions ( @xmath161 can be measured independently from the @xmath620 , see eq .
( [ f30 ] ) and @xmath162 - from the @xmath564 , see eq .
( [ f36 ] ) ) .
therefore , the nonlinear ( in @xmath143 ) resistive behavior of the vortex system can be caused by factors of both an anisotropic and isotropic pinning origin .
it is important to underline that whereas the structure of the @xmath107 and @xmath160 is the same as for purely @xmath64-pinning problem , the structure of the @xmath231 and @xmath322 is strongly different from the structure of the purely @xmath65-pinning problem due to the fact that @xmath322 , as motive force of the ( @xmath63)-problem , is nonlinear and anisotropic ( see eqs .
( [ f34])-([f35 ] ) ) and figs . 3 , 4 , 5 ) .
two main new features appear due to the introduction of the isotropic @xmath65-pins into the initially anisotropic @xmath64-pinning problem .
first , unlike the stochastic model of uniaxial anisotropic pinning studied previously@xmath12 , where the critical current density @xmath53 is indeed equal to zero for all directions ( excepting @xmath45 ) due to the guiding of vortices , in the given @xmath63 model the anisotropic critical current density @xmath58 exists for all directions because @xmath65-pins `` quench '' the guiding of vortices in the limit @xmath621 .
second , the hall resistivity response functions @xmath502 can have a change of sign in a certain range of @xmath485 ( at fixed dimensionless hall constant @xmath622 and the dimensional hall conductivity @xmath623 , whereas the sign of the @xmath492 does not change .
it should be noted that recently@xmath359 the nonlinear ( in @xmath178 ) anisotropic longitudinal and transverse resistances of nb films deposited on facetted sapphire substrates were measured at different angles @xmath3 between @xmath44 and facet ridges in a broad range of temperature and relatively small magnetic field @xmath624 .
the experimental data were in good agreement with the theoretical model described here .
the measured @xmath625 dependences can be fitted using the probability functions @xmath184 and @xmath626 in the form proposed here ( see eq .
( [ f29 ] ) ) with the anisotropic and isotropic pinning potential given by eq .
( [ f28 ] ) .
the periods and depths of the potential wells were estimated from the experimental data@xmath627 and were used here ( see sec.iii ) for the theoretical analysis of different types of nonlinear anisotropic @xmath66-dependent magnetoresistivity responses , given by eqs .
( [ f30])-([f33 ] ) , in the form of graphs ( see figs . 3 - 18 ) .
whether these theoretical results can explain a new portion of the @xmath66-dependent @xmath63 resistivity data measured at @xmath628 ( in particular , for the samples investigated earlier@xmath627 at small current densities ) remains to be seen .
niessen and c.h .
weijsenfeld , j. appl . phys .
* 40 * , 384 ( 1969 ) .
prodan , v.a .
shklovskij , v.v .
chabanenko et al . , physica c * 302 * , 271 ( 1998 ) .
chabanenko , a.a .
prodan , v.a .
shklovskij et al . , physica c * 314 * , 133 ( 1999 ) .
h.pastoriza , s.candia , and g.nieva , phys .
. lett . * 83 * , 1026 ( 1999 ) .
g. danna , v. berseth , l. forro , a. erb , e. walker , phys .
b * 61 * , 4215 ( 2000 ) .
m. huth , k. a. ritley , j. oster et al . ,
funct . mater .
* 12 * , 333 - 341 ( 2002 ) .
o.k.soroka , m.huth , v.a .
shklovskij et al . , physica c * 388 - 389 * , 773,(2003 ) .
o.k.soroka,``vortex dynamics in superconductors in the presence of anisotropic pinning ''
d. thesis , j. gutenberg university , mainz , 2005 .
o.k.soroka , v.a.shklovskij , m.huth et al .
, to be published .
y. mawatari , phys .
b * 56 * , 3433 ( 1997 ) .
shklovskij , a.a .
soroka , a.k .
soroka , zh eksp .
. fiz . * 116 * , 2103 ( 1999 ) [ jetp 89 , 1138 ( 1999 ) ] .
g. blatter , m.v .
feigelman , v.b .
geshkenbein et al . , rev .
phys . * 66 * , 1125 ( 1994 ) .
o.v . usatenko and v.a .
shklovskij , j. phys . a 27 , 5043 ( 1994 ) . v.a .
shklovskij , phys . rev .
b * 65 * , 092508 ( 2002 ) .
shklovskij , j. low temp . phys . * 130 * , 407 ( 2003 ) .
shklovskij , j. low temp . phys .
* 139 * , 289 ( 2005 ) .
vinokur , v.b .
geshkenbein , m.v .
feigelman , and g. blatter , phys .
71 * , 1242 ( 1993 ) .
v.a . shklovskij .
physica c * 388 - 389 * , 655 ( 2003 ) . b. chen and j. dong , phys .
b * 44 * , 10206 ( 1991 )
. v.a . shklovskij and a.a .
soroka , fiz .
nizk . temp .
* 28 * , 365 ( 2002 ) ; [ low temp
* 28 * , 254 ( 2002 ) ] . | explicit current - dependent expressions for anisotropic longitudinal and transverse nonlinear magnetoresistivities are represented and analyzed on the basis of a fokker - planck approach for two - dimensional single - vortex dynamics in a washboard pinning potential in the presence of point - like disorder .
graphical analysis of the resistive responses is presented both in the current - angle coordinates and in the rotating current scheme .
the model describes nonlinear anisotropy effects caused by the competition of point - like ( isotropic ) and anisotropic pinning .
nonlinear guiding effects are discussed and the critical current anisotropy is analyzed . gradually increasing the magnitude of isotropic pinning force
this theory predicts a gradual decrease of the anisotropy of the magnetoresistivities .
the physics of transition from the new scaling relations for anisotropic hall resistance in the absence of point - like pins to the well - known scaling relations for the point - like disorder is elucidated .
this is discussed in terms of a gradual isotropizaton of the guided vortex motion , which is responsible for the existence in a washboard pinning potential of new ( with respect to magnetic field reversal ) hall voltages .
it is shown that whereas the hall conductivity is not changed by pinning , the hall resistivity can change its sign in some current - angle range due to presence of the competition between _ i_- and _ a_-pins . |
information theory usually deals with the transmission of a sequence of discrete symbols , such as 0 and 1 .
even if the information to be transmitted is of continuous nature , such as the position of a particle , it can be represented with arbitrary accuracy by a string of bits . however , there are situations where information can not be encoded in such a way . for example
, the emitter ( conventionally called alice ) wants to indicate to the receiver ( bob ) a direction in space .
if they have a common coordinate system to which they can refer , or if they can create one by observing distant fixed stars , alice simply communicates to bob the components of a unit vector along that direction , or its spherical coordinates @xmath0 and @xmath1 .
but if no common coordinate system has been established , all she can do is to send a real physical object , such as a gyroscope , whose orientation is deemed stable . in the quantum world , the role of the gyroscope is played by a system with large spin .
for example , alice can send angular momentum eigenstates satisfying @xmath2 .
this is essentially the solution proposed by massar and popescu @xcite who took @xmath3 parallel spins , polarized along .
this , however , is not the most efficient procedure : for two spins , a higher accuracy is achieved by preparing them with opposite polarizations @xcite . for more than two spins , optimal results are obtained with entangled states @xcite .
the above discussion can be generalized to the transmission of a cartesian frame .
if @xmath3 spins are available , one can encode a cartesian frame in an entangled state of these spins , as in @xcite .
however , a more accurate transmission is then obtained if alice uses half of the spins to indicate the @xmath4-axis , and the other half for her @xmath5-axis @xcite . in this case
the @xmath4 and @xmath5 directions found by bob may not be exactly perpendicular and some adjustment will be needed to obtain bob s best estimate of the @xmath4 and @xmath5 axes .
finally , the @xmath6-axis can be inferred from the estimates of the @xmath4 and @xmath5 axes . however , it is not possible to proceed in this way if a _
quantum messenger is available .
the optimal transmission of a cartesian frame by a hydrogen atom ( formally , a spinless particle in a coloumb potential ) was derived by peres and scudo @xcite .
the results of @xcite can also be used for a hydrogen atom if one considers the angular momentum eigenstates to be those of the atom , rather than those of @xmath3 spins . in this paper
we show how to transmit a cartesian frame by using the elliptic rydberg states of a hydrogen atom .
these are the quantum mechanical analog of a classical keplerian orbit , and it is known how to produce these states experimentally .
elliptic rydberg states , just as their classical counterparts , define three orthogonal directions in space , and thus are natural candidates for encoding a cartesian frame . in the following section we discuss the properties of quantum elliptic states . section iii deals with the transmission of one direction by means of them , and in sec .
iv we use them to transmit two orthogonal axes ( and thus a cartesian frame ) . in sec .
iii and iv the detection procedure is based on so(3 ) coherent states as in @xcite .
so(4 ) coherent states are employed in sec .
v to produce a positive operator valued measure ( povm ) which enables the use of elliptic states for the transmission of two directions that are not orthogonal .
even when these states are used by alice to transmit two orthogonal axes , the two directions found by bob are not necessarily perpendicular and further adjustment is needed , as explained above .
as shown in the appendix , these adjustments increase the fidelity of transmission of two orthogonal axes .
however , a higher fidelity is achieved with a povm based on the so(3 ) rotation group , especially when the energy quantum number @xmath7 is large .
a classical bounded keplerian orbit in a potential @xmath8 can be defined by its constants of motion : the energy @xmath9 , the angular momentum @xmath10 which is an axial vector perpendicular to the plane of the orbit , and the laplace - runge - lenz ( lrl ) vector @xcite = ( -2h)^-1/2(*p*-k*r*/r ) , where @xmath11 is the particle s reduced mass , and we introduced a prefactor @xmath12 for later convenience .
this prefactor , which is a constant of motion , does not appear in the usual definition of the lrl vector @xcite .
the hamiltonian is @xmath13 .
we consider only bounded motion for which the energy @xmath14 , which is the numerical value of @xmath15 , is negative .
the classical orbit is then an ellipse , and the lrl vector is a polar vector directed along its major axis .
it satisfies : = 0 , [ classical1 ] and ^2 + ^2 = -^2k^2 /2e .
[ classical2 ] because of these relations , only five out of the seven constants are independent .
they uniquely determine the shape and orientation of the ellipse .
its eccentricity is @xcite e = || /k .
[ eccentricity ] we now turn to the quantum version .
we use natural units , @xmath16 , so that the energy levels for bound states are @xmath17 , where @xmath7 is an integer .
the operator @xmath18 is defined by = ( -2h)^-1/2((*p*-*l*)-*r*/r ) .
note that @xmath15 commutes with @xmath19 .
the commutation relations for the operators @xmath20 and @xmath21 are @xcite = i _ ijkk_k , [ com1 ] = i _ ijkl_k . [ com2 ] together with @xmath22=i \epsilon_{ijk}l_{k}$ ] , these are the commutation rules of infinitesimal rotations in four dimensional euclidean space , which leave the @xmath7-th energy level subspace invariant @xcite . the coherent states of so(4 ) , _ i.e. _ the states for which the dispersion of @xmath23 is minimal , can be built from the coherent states of so(3 ) , since so@xmath24so(3)@xmath25so@xmath26 . define _ 1 = ( * l * - * k*)_2=(*l * + * k * ) .
[ j2 ] these two operators have the commutation relations of two independent three - dimensional angular momenta : = i _ ijkj_1k , = i _ ijkj_2k , = 0 . instead of the classical equations ( [ classical1 ] ) and ( [ classical2 ] ) we now have @xcite = * k * = 0 , and ^2 + * k*^2 = -1 -1/2h = n^2 - 1 , [ quantum2 ] where the last form of the equality holds for energy eigenstates . in the classical limit
, @xmath27 , eq . ( [ quantum2 ] ) reduces to the classical one ( [ classical2 ] ) .
in the rest of this article we consider only energy eigenstates . owing to eq .
( [ j2 ] ) , we have j_1(j_1 + 1 ) = j_2(j_2 + 1)= ( n^2 - 1 ) .
[ j1=j2 ] where @xmath28 and @xmath29 are the quantum numbers referring to the operators @xmath30 and @xmath31 respectively .
it follows that @xmath28 and @xmath29 are equal : @xmath32 , and @xmath33 , so that j=(n-1 ) .
the coherent states of a three - dimensional angular momentum will be denoted by @xmath34 .
they obey @xmath35 for an arbitrary classical unit vector @xmath36 .
for the coherent states the dispersion @xmath37 is minimal : @xmath38 .
in particular , @xmath39 , and @xmath40 , where @xmath41 and @xmath42 with @xmath43 .
the coherent states of so(3 ) are obtained by a rotation of a fiducial coherent state @xmath44 , |j , _
= e^-i l_z e^-i l_y |j , * z*. the coherent states of so(4 ) are now obtained as direct products of coherent states for each of the so(3 ) subgroups , |n , * u*_1 * u*_2 = |j_1 , * u*_1 |j_2 , * u*_2 .
[ product ] where the unit vectors @xmath45 and @xmath46 are again classical .
the coherent state @xmath47 obeys _ 1_1|j_1 , _ 1=j|j_1 , _ 1=(n-1)|j_1 , _ 1 , [ defu1 ] and
likewise we have _ 2_2|j_2 , _ 2=j|j_2 , _ 2=(n-1)|j_2 , _ 2 .
[ defu2 ] as from now , we shall omit the symbols @xmath48 and @xmath49 in state vectors , since the quantum numbers @xmath50 have fixed values , related by eq .
( [ j1=j2 ] ) . for example
the state @xmath51 which obeys @xmath52 , @xmath53 , and @xmath54 will be written simply as @xmath55 , @xmath56 becomes @xmath57 and @xmath58 becomes @xmath59 , etc .
the symbol @xmath60 will always denote the fixed value @xmath61 .
owing to eq .
( [ j2 ] ) , the dispersion of @xmath23 is minimal for coherent states @xcite : ( ) ^2 + ( ) ^2 = 2[(_1)^2 + ( _ 2)^2 ] = 2(n-1 ) . to obtain the expansion of the coherent state @xmath62 in the familiar @xmath63 basis , we first expand each of the @xmath64 in eq .
( [ product ] ) : |_i=_m =- j^j d^j_m ( _ i_i ) |j m , i=1,2 , [ ji ] where the @xmath65 are related to the usual rotation matrices @xcite : d^j_m ( ) & & ^(j)(0)_mj , + & = & ^1/2()^j+m ( ) ^j - m e^-im .
substitution into eq .
( [ product ] ) gives m_2 .
[ almosthere ] we then use the angular momentum addition formula , |j m_1|j m_2 = _
l=0 ^ 2j_m =- l^l c^jjl_m_1m_2 m |lm , [ clebsch ] where @xmath66 is the clebsch - gordan coefficient @xcite which vanishes for @xmath67 . combining eqs .
( [ almosthere ] ) and ( [ clebsch ] ) , we finally get d^j_m_2(_2 _ 2)c^jjl_m_1m_2m)|lm .
the classical orbit that corresponds to the coherent state @xmath68 , in the limit of large @xmath7 , can be obtained as follows . from ( [ j2 ] ) , we have = ( * j_1 * + * j_2*)= ( * j_2 * - * j_1 * ) .
[ la ] let @xmath69 be half the angle between @xmath45 and @xmath46 , _
i.e. _ @xmath70 , and define three orthogonal classical unit vectors , * k * , [ kl ] and @xmath71 . denoting by @xmath72 an arbitrary vector orthogonal to @xmath45 , we have _ 1 = ( _ 1_1)_1+(_1_1 ) _ 1 , _ 1_2 = ( _ 1_1)(_1_2)+(_1_1 ) ( _ 1 _ 2 ) .
then from ( _ 1_1)|_1 = j , and _ 1_1|_1=0 , we get ( _ 2_1)|_1&= & _ 2|(_1_2)|_2 , + & = & j_1_2=j2 .
[ melement ] noting that |_1+_2| = 2||,and |_1-_2| = 2|| , [ length ] we obtain from ( [ melement ] ) the expectation values of the components of @xmath21 and @xmath20 along the directions of @xmath73 , @xmath74 and @xmath75 , for the coherent state @xmath76 k_k = ( n-1 ) , [ expk ] l _ = ( n-1 ) , [ expl ] where @xmath77 , etc . in the perpendicular directions
the expectation values vanish : k _ = k_w = l_k = l_w = 0 .
[ experp ] from eqs .
( [ expk])([experp ] ) we see that in the limit of large @xmath7 , the coherent state @xmath68 corresponds to a classical elliptic trajectory in the @xmath78 plane with the lrl vector in the * k * direction , the angular momentum in the @xmath74 direction , and eccentricity @xmath79 , which is the quantum mechanical analog to eq .
( [ eccentricity ] ) .
the unit vector @xmath75 is parallel to the minor axis of the ellipse .
we now turn to the use of elliptic wave functions as direction indicators .
consider two observers ( alice and bob ) who do not have a common reference frame .
alice wants to indicate to bob her @xmath6-axis by using a hydrogen atom in a rydberg state . in the next section
, we shall likewise discuss the transmission of two orthogonal axes by a single hydrogen atom .
we use as much as possible the same notations in both sections .
alice s signal is |a = _ l=0
^ 2j_m =- l^l a_lm |l m , [ alicegeneral ] where _
l=0 ^ 2j_m =- l^l |a_lm|^2 = 1 .
bob s detectors have labels @xmath80 which indicate the unknown euler angles relating his cartesian axes to those used by alice .
the mathematical representation of his apparatus is a povm de()= , where de()=d _ u()|bb|u^(),[povm ] and @xmath81 is the so(3 ) haar measure for euler angles @xcite . as usual
, @xmath82 is the unitary operator for a rotation by euler angles @xmath80 , and @xmath83 is bob s fiducial vector defined as in @xcite , | b = _
l=0 ^ 2j_m =- l^l b_lm|lm , [ bobgeneral ] where for each @xmath84 , _
l^l |b_lm|^2 = 1 .
note that eq .
( [ alicegeneral ] ) was written with alice s notation , while eq .
( [ bobgeneral ] ) is in bob s notation ( recall that they use different coordinate systems ) . optimizing the transmission fidelity , defined by eq .
( [ fidelity ] ) below , leads to @xcite b_lm = a_lm(_n =- l^l|a_ln|^2)^-1/2 , [ bob ] for each @xmath84 .
since @xmath83 is a direct sum of vectors , one for each value of @xmath84 , then likewise @xmath85 is a direct sum with one term for each irreducible representation : u ( ) = _ l^(l ) ( ) , where the @xmath86 are the usual irreducible unitary rotation matrices @xcite .
a generalization of schur s lemma @xcite confirms that eq .
( [ povm ] ) is indeed satisfied , owing to the coefficients @xmath87 in eq .
( [ bobgeneral ] ) .
the fidelity of the transmission of a single direction is defined as usual : f = ^2(/2 ) = ( 1 + ) , [ fidelity ] where @xmath88 is the angle between the direction indicated by alice and the one that is estimated by bob .
if alice indicates her @xmath6-axis , we thus want to maximize @xmath89 .
following the method of peres and scudo @xcite we define euler angles @xmath90 whose effect is rotating bob s cartesian frame into his _ estimate _ of alice s frame , and then rotating back the result by the _ true _
angles from alice s to bob s frame .
the angles @xmath90 thus indicate bob s measurement error . since in this case
bob s estimate refers to alice s @xmath6-axis only , the angle @xmath91 is identical to the second euler angle @xmath92 : _ z = d_| a|u()| b|^2 .
[ expectation ] let us examine two extreme cases .
first , we take alice s vector to be a circular state , with null eccentricity ( @xmath93 _ i.e. _ , |a = | ll , with l = 2j= n-1 .
bob s vector is obtained from ( [ bob ] ) , which in this case gives |
b=|l l. we then have @xcite |u()| b = e^i(n-1)(+ ) ^2(n-1)2 . inserting the last equation into eq .
( [ expectation ] ) gives _
z&=&(n-)_0^ d^4(n-1)(/2 ) , [ wzcircular ] + & = & ( n-1)/n .
infidelity " @xmath94 , whose typical meaning is bob s mean square error @xcite , is ( 1-_z)=1/2n .
[ msecirc ] this result is identical to the one obtained by massar and popescu @xcite , when the number of their parallel spins is taken to be @xmath95 .
the other extreme case corresponds to a classical elliptical orbit with unit eccentricity , so that @xmath96 .
let @xmath18 lie in the @xmath97 direction .
the corresponding quantum state , denoted by @xmath98 , is an extreme stark state with @xmath99 and maximal @xmath100 : |k , * z * @xmath101 and it is also an eigenstate of @xmath102 : @xmath103 owing to ( j_2z - j_1z)|j ,- j |j j =- j_1z|j ,- j j_2z|j j. in the @xmath63 basis we have |k , * z * = _
l=0 ^ 2j_m =- l^lc^jjl_m_1m_2m|lm = _ l=0
^ 2jc^jjl_-j j 0|l0 , [ extreme ] since @xmath104 and @xmath105 .
the fidelity of transmission by this state will be evaluated at the end of this section .
both the circular state and the extreme stark state are coherent states of so(4 ) , but only the circular state is also an angular momentum coherent state .
moreover , the circular state is symmetric , @xmath106 , while the extreme stark state state is not .
this can be seen from @xcite : better results when used by alice to transmit the directions of her @xmath6-axis .
the overlap between two angular momentum coherent states is @xcite |*u*_1|*u*_2|^2= ^4j(/2 ) , [ overlapj ] where @xmath107 is the angle between the directions of @xmath45 and @xmath46 .
it is noteworthy that the overlap between two extreme stark states is the same , as we will see shortly .
first , a rotation of the @xmath108 state by angles @xmath109 gives |k,_= e^-il_z e^-il_y|k , , where again the operator @xmath110 performs an active rotation of the vector @xmath111 . using ( [ la ] )
we have @xmath112 owing to eqs .
( [ com1 ] ) and ( [ com2 ] ) .
thus the rotated extreme stark state is just |k , _
= |-_| _ , where the so(3 ) coherent states @xmath113 are defined as in eq .
( [ ji ] ) .
this stark state is an eigenstate of @xmath114 with the maximal eigenvalue @xmath115 , and it satisfies @xmath116 , as can be seen from eq .
( [ kl ] ) .
the overlap between two such states , @xmath117 , is | -*u*| -*u*|^2| * u*|*u*|^2 , which by using ( [ overlapj ] ) is just : ^4j_1(/2)^4j_2(/2)=^4(n-1)(/2 ) , [ overlapa ] where @xmath107 is the angle between the vectors @xmath118 and @xmath119 . such a simple expression can not hold for the overlap of two generic elliptic states whose eccentricities are not 0 or 1 . let a generic elliptic state |*u*_1 * u*_2 = |*u*_1 |*u*_2 , be an elliptic state with eccentricity @xmath120 .
unlike the @xmath121 and @xmath122 cases , this state does not define one direction , but two independent ones @xmath45 and @xmath46 . if it is rotated by euler angles
@xmath123 the result is e^-il_ze^-il_ye^-il_z|*u*_1**u**_2 = u_1|**u**_1 u_2|**u**_2 . where u_1=e^-ij_1ze^-ij_1ye^-ij_1z , [ uj ] and likewise for @xmath124 .
to obtain this result we have used ( [ la ] ) and the commutation relations ( [ com1 ] ) and ( [ com2 ] ) .
the rotation @xmath125 opens an angle @xmath126 between the classical vectors @xmath127 and @xmath128 , and an angle @xmath129 ( which is generally different from @xmath126 ) between @xmath46 and @xmath130 . here
@xmath131 denotes the classical rotation matrix @xcite .
it follows that _ 1_2|e^-il_ze^-il_y e^-il_z|_1_2= ( ) ^2(n-1).generally , both @xmath126 and @xmath129 are different from the angle between the directions @xmath132 and @xmath133 , or between the directions @xmath74 and @xmath134 .
we now calculate the transmission fidelity for the case where alice sends an extreme stark state @xmath135 .
since @xmath136 contains only @xmath137 terms , so does bob s fiducial vector b_lm = a_l0(|a_l0|^2)^-1/2_m 0 .
we thus have b_lm = _ m 0(a_l 0/|a_l 0| ) , [ bobstark ] | b = _
l=0^n-1(a_l 0/|a_l 0|)|l0 . in order to determine @xmath89 in eq .
( [ expectation ] ) , we note that |u()|b & = & _ l=0^n-1a_l0^*b_l0 l0| ^(l)()|l 0 , + & = & _ l=0^n-1|a_l0|d_00^(l ) ( ) .
we insert this expression into ( [ expectation ] ) .
the result , obtained by using eqs .
( 19)(21 ) of ref .
@xcite , is _
z=_k la_l k|a_l 0a_k 0| , [ fidelm0 ] where @xmath138 is a real symmetric matrix whose non - vanishing elements are a_l , l-1=a_l-1,l = l/ , [ calc4 ] and @xmath139 .
the results are summarized in fig . 1 , in which the mean square error is plotted versus @xmath7 .
the @xmath135 state gives better fidelity than the circular state @xmath140 , but for @xmath141 its fidelity is substantially less than optimal @xcite and goes asymptotically to @xmath142 .
this raises the question whether it is possible to build a
natural " povm by setting bob s vector to @xmath143 , so that povm elements are n|k , * u*_k,*u*_| , where @xmath144 was defined in eq .
( [ starkgeneral ] ) and @xmath3 is a normalization factor .
unfortunately , @xmath111 contains a superposition of all values of @xmath84 , as can be seen from ( [ extreme ] ) .
thus @xmath111 does not belong to one irreducible subspace of the representation of the so(3 ) rotation group .
as a result , the operator b = d_|k , * u*_k , * u*_| is not proportional to the identity , but is a block - diagonal matrix with different blocks for each irreducible representation of the rotation group .
moreover , the resulting povm includes an element which corresponds to the absence of any answer , thus reducing fidelity . a natural povm which uses the so(4 ) group
will be discussed in sec .
the direction of the minor axis of a classical nondegenerate ellipse is that of @xmath145 .
a quantum ellipse also has this property . taking alice s state as a quantum ellipse with eccentricity @xmath120 , with both @xmath73 and @xmath74 lying in the @xmath146 plane so that @xmath147 , the resulting fidelity can be compared with the cases where @xmath73 or @xmath74 points along the @xmath6-axis and the eccentricity of the ellipse is 0 or 1 , respectively .
the fidelity for transmission using the semi - minor axis reaches a maximum at eccentricity of about @xmath148 ( a different eccentricity for each value of @xmath7 ) .
a comparison of the mean square error for using the three options is given in table 1 .
we shall now compare the extreme stark state with alice s optimal vector for the transmission of one axis as calculated in @xcite .
they are both eigenstates of @xmath149 with @xmath137 , and since eq .
( [ fidelm0 ] ) holds , we will present them in the notation : @xmath150 , where @xmath151 as before .
for @xmath152 we have |k , = ( , , ) , while alice s optimal state is |
a_opt = ( , , ) .
thus for @xmath152 the overlap between the extreme stark state and the optimal state is |k , | a_opt|^2 = 0.993491 .
both states give almost the same fidelity for transmission of one axis . for higher values of @xmath7 , they become more and more different . for @xmath153
the overlap is |k , | a_opt|^2 = 0.76406 .
the various components are given in table 2 .
we see that the extreme stark state has coefficients peaked at lower values of @xmath84 than the optimal state .
alice now wants to transmit a cartesian frame by indicating the directions of two axes , the third one being inferred from them . which elliptic state is optimal ? obviously , states with @xmath122 and @xmath121 will not do in this case , since they define only one direction .
we have to find the optimal eccentricity .
let k_= , l_= , where k_= = , owing to eq .
( [ experp ] ) .
we define similar expressions for the other components .
when we want to transmit @xmath73 , namely the direction of the classical lrl vector , then a smaller @xmath154 improves the fidelity .
a similar argument holds for the transmission of @xmath74 .
thus when transmitting two axes , a heuristic guideline is to look for states that satisfy .
a straightforward calculation @xcite gives = l_w = , = , = .
together with eqs .
( [ expk ] ) and ( [ expl ] ) , this gives an equation for the eccentricity , .
therefore we expect that the optimal eccentricity is approximately e===1/. [ opte ] more accurate numerical results are given below .
we now evaluate the fidelity for the transmission of two axes .
alice uses an elliptic state with @xmath155 and @xmath156 ( the unit vectors in the @xmath4 and @xmath5 directions respectively ) .
the eccentricity @xmath157 has to be optimized .
recall the @xmath69 is defined to be half the angle between @xmath45 and @xmath46 .
the definitions of @xmath73 and @xmath74 are given in eq .
( [ kl ] ) . thus in order to meet the above requirements we set in @xmath158 the parameters @xmath159 , and _ 1=- ,
[ alice2 ] fidelities now must be defined for each one of the axes . note that @xmath160 ( for the _ k_th axis ) is given by the corresponding diagonal element of the orthogonal ( classical ) rotation matrix .
for the transmission of the @xmath4 and @xmath5 axes , we thus need @xcite _ x + _ y = ( 1 + ) ( ( + ) .
we expand @xmath161 and @xmath162 as in ( [ alicegeneral ] ) and ( [ bobgeneral ] ) .
bob s optimal fiducial vector is still given by ( [ bob ] ) , and @xmath163 is calculated using equations ( 23)(26 ) of ref .
@xcite . the mean square error per axis
is plotted in fig . 3 for @xmath164 and @xmath165 .
the error is minimal at @xmath166 for @xmath167 , at @xmath168 for @xmath153 , and @xmath169 for @xmath170 .
note that the shape of the curve flattens with increasing @xmath7 , so that the minimum is hard to find numerically .
the intuitive explanation is that in the limit of large @xmath7 , as if alice were to transmit a classical atom , " _
i.e. _ , a classical two body kepler system , then the direction of the classical angular momentum and lrl vectors could be found irrespective of the eccentricity .
therefore , the transmission accuracy would be the same for any eccentricity that is not close to zero or one .
the deviation of the optimum from @xmath171 was expected , since transmission of the @xmath73 direction ( @xmath121 ) achieved higher fidelity than the transmission of the @xmath74 direction ( @xmath122 ) .
thus the ellipse with optimal eccentricity for transmission of two axes is biased to give @xmath172 in order to compensate the difference and make the contribution to the error from the @xmath73 direction about equal to that from the @xmath74 direction .
elliptic states give results very close to the optimal ones . the mean square error for transmission of two axes by elliptic states with optimal eccentricity
is compared to the optimal results @xcite in table iii .
we now construct a povm based on the so(4 ) group and use it in order to transmit two axes .
this povm is naturally built with the so(4 ) coherent states which are , as we have seen , direct products of two so(3 ) coherent states .
we shall use for each one of the so(3 ) subspaces the notation to be transmitted , and @xmath173 as in eq .
( [ povm ] ) . by applying schur s lemma to each of the so(3 )
subspaces we have de_1(_1_1_1)de_2(_2_2_2)=_1_2=. we are now ready to discuss the transmission of alice s @xmath4 and @xmath5 axes by means of an elliptic state .
we take |a = | = define a fiducial vector for bob |b = ( 2j+1)| , [ bobso4 ] written in bob s notations .
thus the povm element is constructed from the vector the result of bob s measurement consists of two sets of euler angles , @xmath174 and @xmath175 .
the first one gives bob s estimate of the active rotation needed to bring his @xmath4-axis to alice s @xmath4-axis .
likewise , the second set gives bob s estimate of the active rotation needed to bring his @xmath5-axis to alice s @xmath5-axis .
the detection probability of these sets of angles is dp(_1 ...
_2)= d__1_1_1d__2_2_2 |a|u_1 u_2|b|^2 . [ prob2notation ] recall that eq .
( [ aliceso4 ] ) was written in alice s notations , while ( [ bobso4 ] ) is in bob s notations . to compute the result explicitly , we need a uniform system of notations . for this
we introduce , as in @xcite , the euler angles @xmath176 that rotate bob s @xmath177 axes into alice s axes .
( the euler angle @xmath69 should not be confused with the eccentricity parameter introduced in sec .
the unitary operator @xmath178 represents an active transformation of bob s state vectors to the corresponding state vectors of alice s system .
therefore , @xmath179 is also the passive transformation from alice s notations to bob s notations . written in bob s notations , alice s vector @xmath180 becomes @xmath181 so that in eq .
( [ prob2notation ] ) , @xmath182 becomes @xmath183 .
owing to the commutation relations ( [ com1 ] ) and ( [ com2 ] ) u()&= & e^-il_ze^-il_ye^-il_z , + & = & u_1()u_2 ( ) , where again @xmath184 and @xmath124 are defined as in eq .
( [ uj ] ) .
we thus have u()|_1_2=u_1()|_1u_2()|_2 .
let us therefore define u_1(_1 _ 1 _ 1 ) = u_1^ ( ) u_1(_1_1_1 ) , [ u1 ] and u_2(_2 _ 2 _ 2 ) = u_2^()u_2(_2_2_2 ) .
[ u2 ] we shall henceforth use the left hand sides of eqs .
( [ u1 ] ) and ( [ u2 ] ) as the new definitions of the symbols @xmath184 and @xmath124 .
as before , the euler angles @xmath185 have the effect of rotating bob s @xmath4-axis into his estimate of alice s @xmath4-axis and then rotating back the result by the true rotation from alice s to bob s frame .
the action of the euler angles @xmath186 is similar for the @xmath5-axis .
thus the euler angles @xmath187 indicate bob s measurement error , and the probability of that error is dp(_1 ...
_2)= d__1 _ 1 _ 1d__2 _ 2 _ 2| a|u_1u_2|b|^2 .
[ prob ] note the similarity with eq .
( [ prob2notation ] ) .
the difference is that ( [ prob2notation ] ) referred to the probability of _ detection _ of a particular set of euler angles , while ( [ prob ] ) gives the probability of _ error _ in that detection .
the transmission mean square error _ per axis _ is , as in eq .
( [ msecirc ] ) , r = ( 1-_x)+(1-_y ) , where @xmath188 and @xmath189 are the angles between the true and estimated directions of the @xmath4-axis and @xmath5-axis , respectively .
since bob infers the direction of the @xmath4-axis from the angles @xmath174 , the value of @xmath190 depends only on @xmath191 .
likewise , the value of @xmath192 depends only on the angles @xmath186 .
we have _ x = d__1 _ 1 _ 1 | | u_1||^2_x , [ wxso4 ] where we have used ( [ prob ] ) and schur s lemma for the second set of angles , namely ( 2j+1)d__2 _ 2 _ 2 u_2||u_2^= _ 2 .
the evaluation of eq .
( [ wxso4 ] ) is identical to the one performed in eq .
( [ wzcircular ] ) , with @xmath7 replaced by @xmath193 everywhere , and we get _ x = ( n-1)/(n+1 ) . likewise _
y = ( n-1)/(n+1 ) .
thus the infidelity ( mean square error ) per axis is ( 1-_x)+(1-_y ) = 1/(n+1 ) in ref .
@xcite it was found that the _ optimal _ povm ( not restricted to elliptic states ) for transmission of two axes using a hydrogen atom is of the form given by eq .
( [ povm ] ) .
it was shown that using this povm , the infidelity per axis for alice s optimal signal approaches @xmath194 asymptotically . using so(4 ) instead of so(3 ) as in @xcite , we obtain an infidelity per axis that is exactly @xmath195 for all values of @xmath7 . as shown in the appendix , an adjustment procedure to obtain to orthogonal axes will further decrease the mean square error by a factor which for large values of @xmath7 tends to @xmath196 .
the so(4 ) povm also enables the transmission of two directions which are not orthogonal , by means of a single hydrogen atom in an elliptic state . to transmit the directions of two general unit vectors
@xmath197 and @xmath198 , alice s prepares the elliptic state |a = |_1_2 = |_1|_2 , ( in her notations ) while bob s vector is ( in his notations ) |b = ( 2j+1)|_1 .
as before , the infidelity for each direction is @xmath195 .
it should be noted that transmission of two non - orthogonal directions with one hydrogen atom is not possible with the so(3 ) povm .
we have shown how elliptic rydberg states can transfer information on the orientation of one direction , or more generally that of a cartesian frame . for increasing values of @xmath7 , the fidelity obtained for a single direction falls rapidly below the optimal ones . however , for a cartesian frame the results are very close to the optimal ones .
furthermore , elliptic states have the advantage of being experimentally accessible , while preparation of the optimal states seems much more difficult .
note that we have assumed alice and bob have the same chirality .
if their chiralities are opposite , then when angular momenta are used for the transmission , the direction inferred by bob should be reversed ( because directions are polar vectors while angular momentum is an axial vector ) .
however , the lrl vector is also a polar vector , thus even if bob and alice have opposite chiralities , the direction inferred by bob is correct . we have also shown how elliptic rydberg states can be prepared to encode two arbitrary directions , when the measurement is based on the so(4 ) rotation group .
work by ap was supported by the gerard swope fund and the fund for promotion of research .
work by nhl was supported by a grant from the technion graduate school .
as we have seen in sec .
v , bob s estimates of alice s @xmath4 and @xmath5 axes may not be exactly orthogonal .
the probability for the estimate of the @xmath4 axis to have an angular error @xmath188 , as can be seen from eq .
( [ overlapj ] ) , is ( _ x)^2n-2(_x/2 ) , [ probomega ] and likewise for the @xmath5 axis .
thus for large values of @xmath7 , the error probability distribution will be highly peaked .
we now calculate the gain in fidelity achieved if bob performs a simple orthogonalization of his two estimates @xmath199 and @xmath200 , by rotating the two vectors in their plane by the same angle , so that they become orthogonal .
let us define two pairs of spherical angles that give the position of the estimated directions with respect to the ( unknown ) true axes .
these positions are given by _
x=(_1_1 , _ 1_1 , _ 1 ) , [ r1 ] and _ y=(_2_2 , _ 2_2 , _ 2 ) .
[ r2 ] the probability distributions will be denoted @xmath201 . in the limit of large @xmath7 , the deviation angles @xmath188 and @xmath189 are small .
hence the distribution are centered as _
x=(_1-,_1),_y=(_2-,_2- ) , where @xmath202 is peaked around @xmath203 . here
we used the fact that the so(4 ) povm gives probabilities of error for each axis which are identical and independent .
define new variables _ i=_i- , [ tal ] _ 2=_2- .
[ phitilde ] the deviation angles are given by @xmath204 and @xmath205 , namely _
x=_1_11-_1 ^ 2-_1 ^ 2 , [ coswx ] and _
y=_2_21-_2 ^ 2-_2 ^ 2 . let @xmath206 denote the infidelity per axis before the adjustment . the infidelities for both axes are equal , thus + & & ( _ 1 ^ 2+_1 ^ 2 ) d_i_1 ^ 2+_1 ^ 2 , [ fx ] where d_i=(_i,_i)_id_id_i , fullfills d_i=1 .
equivalently , we can write the infidelity in terms of @xmath207 and @xmath208 as g(_2 ^ 2+_2 ^ 2 ) d_i_2 ^ 2+_2 ^ 2 .
[ fy ] in _ first _ order we have , by combining ( [ r1 ] ) and ( [ r2 ] ) with the definitions ( [ tal ] ) and ( [ phitilde ] ) , _
x(1,_1,-_1),_y(-_2,1,-_2 ) , [ rapprox ] and the angle @xmath209 between them is given by = _
x_y_1-_2 [ omega ] .
the bisector of @xmath199 and @xmath200 is given by the _ unit _ vector @xmath210 .
using ( [ rapprox ] ) and keeping only first order terms we have / , where we used |_1+_2|(1+_1-_2 ) .
we can also express the bisector @xmath211 in terms of its spherical angles which we shall denote by @xmath212 . since the errors are small , we have @xmath213 , and it is convenient to define = - .
comparison of the two expressions for @xmath211 gives = + ( _ 1+_2),=(_1+_2 ) .
in first order , as eq .
( [ omega ] ) shows , the orthogonalization consists in changing the angles @xmath214 irrespective of @xmath215 , without changing the @xmath215 themselves .
hence , in first order , the procedure defines _1=- , _2=+ , _1=_2==(_1+_2 ) , where again @xmath216 .
the change in @xmath217 is of higher order , @xmath218 . the new infidelity per axis @xmath219 is g^new=_1^2+_1 ^ 2=_2^2+_2 ^ 2 .
returning to ( [ fx ] ) and ( [ fy ] ) , consider the integrals over @xmath214 .
( ) ( , ) dkeeping in mind that the distributions for @xmath220 and @xmath221 are identical , the @xmath1-part of the infidelity per axis _ before _ the adjustment is [ equal ] + & & = _ 1 ^ 2_(_1 ) _ ( _ 2)d_1d_2/4 , + & & = _ 2 ^ 2_(_2 ) _ ( _ 2)d_1d_2/4 .
the @xmath1-parts of the infidelities _
after _ the adjustment , denoted by @xmath222 , are g^new_= _ 1^2=_2^2=
_ 1 ^ 2 + 2_1_2 + _ 2 ^ 2 .
the functions @xmath223 and @xmath224 are even , because the probability distribution @xmath225 depends only on the angles @xmath188 or @xmath189 , which are independent of the sign of @xmath220 and @xmath221 .
thus _ 1_2=_1_2 _ ( _ 1)_(_2)d_1 d_2=0 . with the help of eq .
( [ equal ] ) we obtain g^new_= _ 1 ^ 2 + _ 2 ^ 2 = _ 1 ^ 2 .
thus the @xmath1-parts of the infidelity are halved , g^new_=g_. as already stated , the angles @xmath226 are unchanged in first order . since the probability function @xmath227 depends only on the angle @xmath228 , it is symmetric with respect to rotations around the @xmath4 axis .
a similar argument holds for the @xmath5-axis .
1 ^ 2=_1 ^ 2=_2 ^ 2=_2 ^ 2 , and we have finally g^new = g .
h. goldstein , _ classical mechanics _ ( addison - wesley , reading , 1980 ) .
landau and e.m .
lifshitz , _ quantum mechanics _ ( pergamon , oxford , 1977 ) .
m. bander and c. itzykson , rev .
phys . * 38 * , 330 ( 1996 ) .
j. c. gay , d. delande and a. brommier , phys .
a * 39 * , 6587 ( 1989 ) .
a. r. edmonds , _ angular momentum in quantum mechanics _ ( princeton university press , princeton , 1957 ) .
m. tinkham , _ group theory and quantum mechanics _ ( mcgraw - hill , new - york , 1964 ) , p. 121 .
e. wigner , _ group theory _
( academic press , new york , 1959 ) , p. 76 . .eccentricities @xmath229 and mean square errors @xmath230 for transmission of a single direction using @xmath231 , @xmath232 , or @xmath233 , for @xmath167 or @xmath234 . [ cols="^,<,<,<",options="header " , ] | the orientation in space of a cartesian coordinate system can be indicated by the two vectorial constants of motion of a classical keplerian orbit : the angular momentum and the laplace - runge - lenz vector . in quantum mechanics ,
the states of a hydrogen atom that mimic classical elliptic orbits are the coherent states of the so(4 ) rotation group .
it is known how to produce these states experimentally .
they have minimal dispersions of the two conserved vectors and can be used as direction indicators .
we compare the fidelity of this transmission method with that of the idealized optimal method . |
* expected order parameter for two incoherent oscillators . * in fig.3 of the main text we showed that when @xmath11 approaches @xmath113 the order parameter @xmath67 for a pair of symmetric oscillators remains higher than the expected value @xmath114 for a pair of incoherent oscillators .
we report here the derivation of @xmath114 for a pair of oscillators that are not synchronized .
we assume that the phases @xmath115 and @xmath116 of the two oscillators are drawn uniformly in @xmath117 $ ] .
thanks to the rotational symmetry , we can set one of the phases equal to @xmath55 , e.g. @xmath118 , while drawing the other uniformly in @xmath117 $ ] .
we notice that the value of @xmath67 for the generic pair of phases @xmath119 reads : @xmath120 consequently , the expected value of @xmath67 is obtained as the average over all the possible choices of @xmath116 , namely : @xmath121 * phase dispersion and numerical estimate for incoherent oscillators . * in fig.2d of the main text we reported , as a function of @xmath11 , the dispersion of the phases @xmath34 for a system of seven oscillators coupled through graph @xmath19 in fig.1a .
we noticed that the dispersion approaches the value @xmath122 when @xmath123 .
given a set of @xmath0 unitary vectors having phases @xmath124 , consider the average vector : @xmath125 having polar coordinates @xmath126 .
the dispersion of the @xmath0 phases of the set around the phase @xmath127 of the average vector is defined as : @xmath128 where the difference @xmath129 ) is computed @xmath130 and takes in @xmath131 $ ] .
an estimate of the expected phase dispersion for a set of @xmath0 incoherent oscillators can be obtained as the average over @xmath132 independent realizations of @xmath34 computed on a set of @xmath0 phases uniformly drawn in @xmath117 $ ] . for the system of @xmath133 oscillators
coupled through graph @xmath19 we averaged @xmath34 over @xmath134 samples , obtaining the estimate @xmath135 .
* computation of the maximum lyapunov exponent @xmath35 .
* the computation of the maximum lyapunov exponent for the system coupled through graph @xmath19 of fig .
( 1 ) in the main text was performed using equally separated values of @xmath11 between @xmath55 and @xmath136 ( @xmath137 ) .
then , for each value of @xmath11 , we considered 500 different initial configurations of the phases of the seven oscillators .
for each initial condition , we let evolve the trajectory @xmath138 according to eq .
( 1 ) in the main text , until it reached the attractor ( or the stationary state , for @xmath139 ) . to properly take into account the rotational symmetry of the system
, we studied the evolution of @xmath138 in cartesian coordinates , i.e. , by looking at the set of @xmath140 variables @xmath141 .
we considered a perturbation of @xmath142 of magnitude @xmath143 , i.e. , a trajectory @xmath144 such that @xmath145 . here
@xmath146 denotes the euclidean distance in @xmath147 .
then , we integrated both trajectories for one integration step @xmath148 ( using a standard fourth - order runge - kutta integration scheme ) , we measured the distance @xmath149 and we computed @xmath150 .
the quantity @xmath151 is a one - step approximation of the largest lyapunov exponent of the system @xcite .
then , we realigned the perturbed trajectory @xmath144 so that the distance between @xmath152 and the realigned perturbed trajectory @xmath153 was equal to @xmath154 in the same direction of @xmath155 , and we iterated the procedure .
the value of @xmath35 for a set of initial conditions was obtained by averaging the values of @xmath151 computed at each iteration over @xmath156 subsequent integration steps .
* brain data acquisition and pre - processing .
* the anatomical connectivity network is based on the connectivity matrix obtained by diffusion magnetic resonance imaging ( dw - mri ) data from 20 healthy participants , as described in @xcite .
the elements of this matrix represent the probabilities of connection between the 90 anatomical regions of interest ( @xmath107 nodes in the network ) of the tzourio - mazoyer brain atlas @xcite .
these probabilities are proportional to the density of fibers between different areas , so each element of the matrix represents an approximation of the connection strength between the corresponding pair of brain regions .
the functional brain connectivity was extracted from bold fmri resting state recordings obtained as described in @xcite .
all fmri data sets ( segments of 5 minutes recorded from 15 healthy subjects ) were co - registered to the anatomical data set and normalized to the standard mni ( montreal neurological institute ) template image , to allow comparisons between subjects . as for dw - mri data , normalized and corrected functional scans were sub - sampled to the anatomical labeled template of the human brain @xcite .
regional time series were estimated for each individual by averaging the fmri time series over all voxels in each region ( data were not spatially smoothed before regional parcellation ) . to eliminate low frequency noise ( e.g. slow scanner drifts ) and higher frequency artifacts from cardiac and respiratory oscillations ,
time - series were digitally filtered with a finite impulse response ( fir ) filter with zero - phase distortion ( bandwidth @xmath157 hz ) as in @xcite . *
functional synchrony . * a functional link between two time series @xmath158 and @xmath159 ( normalized to zero mean and unit variance ) was defined by means of the linear cross - correlation coefficient computed as @xmath160 , where @xmath161 denotes the temporal average . for the sake of simplicity
, we only considered here correlations at lag zero . to determine the probability that correlation values are significantly higher than what is expected from independent time series , @xmath162 values ( denoted @xmath163 )
were firstly transformed by the fisher s z transform @xmath164 under the hypothesis of independence , @xmath165 has a normal distribution with expected value 0 and variance @xmath166 , where @xmath167 is the effective number of degrees of freedom @xcite . if the time series consist of independent measurements , @xmath167 simply equals the sample size , @xmath0
nevertheless , autocorrelated time series do not meet the assumption of independence required by the standard significance test , yielding a greater type i error @xcite .
in presence of auto - correlated time series @xmath167 must be corrected by the following approximation : @xmath168 where @xmath169 is the autocorrelation of signal @xmath170 at lag @xmath171 . to estimate a threshold for statistically significant correlations , a correction for multiple
testing was used .
the false discovery rate ( fdr ) method was applied to each matrix of @xmath165 values @xcite . with this approach ,
the threshold of significance @xmath172 was set such that the expected fraction of false positives is restricted to @xmath173 .
* clustering of phase values . * to identify brain areas that could be related by a topological symmetry , we used the anatomical connectivity obtained from the dw - mri data as the connectivity matrix ( @xmath107 nodes ) in eq .
( 1 ) of the main text .
a standard hierarchical agglomerative clustering algorithm was then used to identify nodes with similar phases @xcite .
the resulting dendrogram is depicted in fig .
[ figure1suppmat ] .
a problem is in the complexity class np if it can be solved in polynomial time by a nondeterministic turing machine .
a problem is said to be np - complete if it is in np and it is np - hard , i.e. , at least as hard to solve as any other problem in np . see ref .
@xcite for an in - depth discussion of complexity classes .
_ computational complexity _ ( addison wesley , reading , ma , 1994 ) . each row and each column of a permutation matrix has exactly one entry equal to one and all others equal to @xmath55 . if @xmath61 is a permutation matrix , the matrix product @xmath176 swaps pairs of rows of @xmath62 , while @xmath177 swaps pairs of columns of @xmath62 . | we study a kuramoto model in which the oscillators are associated with the nodes of a complex network and the interactions include a phase frustration , thus preventing full synchronization .
the system organizes into a regime of remote synchronization where pairs of nodes with the same network symmetry are fully synchronized , despite their distance on the graph .
we provide analytical arguments to explain this result and we show how the frustration parameter affects the distribution of phases .
an application to brain networks suggests that anatomical symmetry plays a role in neural synchronization by determining correlated functional modules across distant locations .
synchronization of coupled dynamical units is a ubiquitous phenomenon in nature @xcite .
remarkable examples include phase locking in laser arrays , rhythms of flashing fireflies , wave propagation in the heart , and also normal and abnormal correlations in the activity of different regions of the human brain @xcite . in 1975
y. kuramoto proposed a simple microscopic model to study collective behaviors in large populations of interacting elements @xcite . in its original formulation
the kuramoto model describes each unit of the system as an oscillator which continuously readjusts its frequency in order to minimize the difference between its phase and the phase of all the other oscillators .
this model has shown very successful in understanding the spontaneous emergence of synchronization and , over the years , many variations have been considered @xcite .
recently , the kuramoto model has been also extended to sets of oscillators coupled through complex networks @xcite , and it has been found that the topology of the interaction network has a fundamental role in the emergence and stability of synchronized states @xcite . in particular , the presence of communities groups of tightly connected nodes has a relevant impact on the path to synchronization @xcite , and units that are close to each other on the network , or belong to the same module or community @xcite , have a higher chance to exhibit similar dynamics .
this implies that nodes in the same structural module share similar functions , which is a belief often supported by empirical findings @xcite .
however , various examples are found in nature where functional similarity is instead associated with morphological symmetry . in these cases , units with similar roles , which could potentially swap their position without altering the overall functioning of the system ,
appear in remote locations of the network .
some examples include cortical areas in brains @xcite , symmetric organs in plants and vertebrates @xcite , and even atoms in complex molecules @xcite .
therefore , identifying the sets of symmetric units of a complex system might be helpful to understand its organization .
finding the global symmetries in a graph , i.e. , constructing its automorphism group , is a classical problem in graph theory .
however , it is still unknown if this problem is polynomial or np - complete @xcite , even if there exist polynomial - time algorithms for graphs with bounded maximum degree @xcite .
recent works have focused instead on defining and detecting local symmetries in complex networks @xcite .
nevertheless , the interplay between the structural symmetries of a network and the dynamics of processes occurring over the network has been studied only marginally @xcite , or for specific small network motifs @xcite . in this letter
we show that network symmetries play a central role in the synchronization of a system .
we consider networks of identical kuramoto oscillators , in which a phase frustration parameter forces connected nodes to maintain a finite phase difference , thus hindering the attainment of full synchronization .
we prove that the configuration of phases at the synchronized state reflects the symmetries of the underlying coupling network .
in particular , two nodes with the same symmetry have identical phases , i.e. , are fully synchronized , despite the distance between the two nodes on the graph .
such a remote synchronization behavior is here induced by the network symmetries and not by an initial _ ad hoc _ choice of different natural frequencies @xcite .
let us consider @xmath0 identical oscillators associated to the nodes of a connected graph @xmath1 , with @xmath2 nodes and @xmath3 links .
each node @xmath4 is characterized , at time @xmath5 , by a phase @xmath6 whose time evolution is governed by the equation @xmath7 here @xmath8 is the natural frequency , identical for all the oscillators , and @xmath9 is the adjacency matrix of the coupling graph .
the model has two control parameters : @xmath10 accounting for the strength of the interaction , and @xmath11 , the phase frustration parameter ranging in @xmath12 $ ] .
when @xmath13 , the model reduces to a network of identical kuramoto oscillators . in this case
, the fully synchronized state is globally stable for a set of initial conditions having finite measure @xcite and the transient dynamics closely reflects the structure of the graph , so that nodes belonging to the same structural module evolve similarly in time @xcite .
however , the synchronized state can coexist with other nontrivial attractors , e.g. , uniformly twisted waves , especially if the coupling topology is regular and sparse ( see ref .
@xcite for a discussion about the size of the sync basin ) . instead ,
if the oscillators are not identical the frequency distribution tends to separate their phases and , as a result , there is a transition from an incoherent state ( with order parameter @xmath14 equal to 0 ) to a synchronized one ( @xmath15 ) at a critical value @xmath16 of the coupling strength . the introduction of a phase frustration @xmath17 forces directly connected oscillators to maintain a constant phase difference @xcite .
in particular , we found that for a wide range of @xmath18 the dynamics in eqs .
( [ kura ] ) reaches a stationary state in which the oscillators at two symmetric nodes have exactly the same phase , and this phase differs from the phases of nodes with different symmetries . let us first illustrate this behavior and the effect of @xmath11 on the three graphs @xmath19 , @xmath20 and @xmath21 shown in fig .
[ fig1 ] . in the three topmost panels of fig .
[ fig2 ] we report the results of the numerical integration of eqs .
( [ kura ] ) on the graph @xmath19 for three different values of @xmath11 .
we find that , after a transient , the system settles into a stationary state in which , at any time @xmath5 , the phases are grouped into four different trajectories : @xmath22 , @xmath23 , @xmath24 and @xmath25 . in general , by increasing @xmath11 up to a certain value @xmath26 we better separate the four trajectories .
the four clusters of nodes obtained for @xmath27 are identified by a color code in fig .
[ fig1 ] .
we notice that each cluster groups together all the nodes with the same symmetry . in this way
two distant nodes of the graph , e.g. node @xmath28 and node @xmath29 , are fully synchronized even if the other nodes in the paths connecting them have different phases . in this respect ,
what we observe is a remote synchronization @xcite .
we have found similar results for the linear chain and for the bethe lattice ( see nodes with the same colors in @xmath20 and @xmath21 in fig .
[ fig1 ] ) . in fig .
[ fig1]a .
panels ( a)(c ) : after an initial transient the system reaches a phase - locked synchronized state in which symmetric nodes have the same phase .
the panels correspond to three different values of the frustration parameter , respectively , ( a ) @xmath30 , ( b ) @xmath31 , ( c ) @xmath32 .
panel ( d ) : for @xmath33 the synchronized state becomes unstable , the order parameter decreases while the dispersion of the phases @xmath34 increases .
panel ( e ) : the maximum lyapunov exponent @xmath35 of oscillators coupled through @xmath19 becomes positive for @xmath33 , and the systems enters a chaotic regime .
the dashed yellow line indicates the approximate position of @xmath26.,width=288 ] notice that if the system reaches a synchronized state and @xmath11 is small enough , eq . ( [ kura ] ) can be linearized by replacing the sinus with its argument .
we obtain @xmath36 ~~ \label{kura_lin}\ ] ] where @xmath37 is the degree of node @xmath4 , @xmath38 are the entries of the laplacian matrix of the graph @xmath39 , and @xmath40 is a diagonal matrix such that @xmath41 . without loss of generality , we can set @xmath42 , @xmath43 . if the system is synchronized then @xmath44 , so that the phases must satisfy the equations @xmath45 $ ] at any time , or equivalently @xmath46 \label{kura_lin_equi}\ ] ] where @xmath47 is the average degree of the network .
this corresponds to a synchronization frequency @xmath48 . in a connected graph
the laplacian matrix has one null eigenvalue and the system of eqs .
( [ kura_lin_equi ] ) is singular .
consequently , at each time @xmath5 we can solve the system by computing the phase difference between each node and a given node chosen as reference .
for instance , if in @xmath19 we define @xmath49 , by solving eqs .
( [ kura_lin_equi ] ) we obtain @xmath50 $ ] , @xmath51 $ ] , and @xmath52 $ ] .
this is in agreement with the results of the simulations : the phases are clustered into four groups , with nodes with the same symmetry having the same phase , and nodes with different symmetries being separated by a phase lag that depends on @xmath11 as in the relations found above .
an analogous analytical expression can be derived for a finite chain ( graph @xmath20 in fig.[fig1 ] ) , for which we obtain @xmath53\alpha $ ] and @xmath54\alpha$ ] .
consequently , two nodes symmetrically placed with respect to node @xmath55 will have identical phases .
we now provide a general argument to explain why the synchronization of eqs .
( [ kura ] ) is related to graph symmetries .
a graph @xmath56 has a symmetry if and only if it is possible to find a bijection @xmath57 which preserves the adjacency relation of @xmath58 , i.e. , which is an automorphism for @xmath58 .
formally , this means that there exists a permutation matrix @xmath59 such that @xmath60 .
if @xmath61 corresponds to an automorphism of @xmath58 then @xmath61 commutes with @xmath62 , i.e. , @xmath63 , and @xmath64 performs a relabeling of the nodes of the original graph which preserves the adjacency matrix @xcite . in general a graph can admit more than one automorphism .
for instance , graph @xmath19 in fig.[fig1 ] has at least three nontrivial bijections which preserve the adjacency matrix , namely @xmath65 when @xmath33 .
panel ( a)(d ) : running averages of @xmath66 ( orange ) and pairwise order parameters @xmath67 ( black , red and green lines ) for typical trajectories of oscillators coupled through graph @xmath19 , when @xmath11 is respectively equal to ( a ) @xmath68 , ( b ) @xmath69 , ( c ) @xmath70 and ( d ) @xmath71 .
the dashed lines indicate the expected synchronization level for a system of two ( blue line , @xmath72 ) and seven incoherent oscillators ( gray line , @xmath73 ) .
panel ( e ) the plot of the phases of pairs of symmetric nodes for @xmath74 in two different temporal intervals ( the shaded gray regions in panel b ) reveal the existence of metastable , partially synchronized states.,width=297 ] node @xmath75 and node @xmath76 are symmetric because we can relabel the nodes of @xmath19 ( e.g. by means of either @xmath77 or @xmath78 ) so that node @xmath75 is mapped into node @xmath76 and vice versa , and the adjacency matrix of @xmath19 is left unchanged .
similarly , for the pairs @xmath79 and @xmath80 , there are two different relabelings which preserve adjacency relations , i.e. , @xmath81 and @xmath78 . in terms of symmetries , the graph @xmath58 has the following four different classes of nodes : @xmath82 , @xmath83 , @xmath84 , @xmath85 .
now , if a permutation of the nodes is an automorphism of @xmath58 , then @xmath86 , i.e. , the associated permutation matrix @xmath61 also commutes with the laplacian matrix of the graph . by
left - multiplying both sides of eq .
( [ kura_lin_equi ] ) by @xmath61 , we get @xmath87 $ ] . since @xmath88 ( @xmath61 commutes with @xmath89 ) and @xmath90 ( symmetric nodes have the same degree ) then we have @xmath91 \label{eq : final}\ ] ] combining eq .
( [ kura_lin_equi ] ) and eq .
( [ eq : final ] ) we finally obtain the linear system @xmath92 which is singular , i.e. , has one free variable .
again , it can be solved by leaving free one of the @xmath0 variables @xmath93 , setting @xmath94 and considering the new system @xmath95 .
the matrix @xmath96 is obtained from @xmath61 by removing the row and the column corresponding to node @xmath4 . if @xmath61 does not permute node @xmath4 with another node , then @xmath96 is still a permutation matrix .
similarly , @xmath97 is the reduced laplacian , i.e. , the matrix obtained from the laplacian by deleting the @xmath98th row and the @xmath98th column . by left - multiplying by @xmath99 , which is not singular ,
we obtain @xmath100 since @xmath101 is a permutation of the phases of symmetric nodes , eq.([eq : phase_perm ] ) implies that the phases of symmetric nodes will be equal at any time , whereas by solving eq .
( [ eq : final ] ) we can get the values of the corresponding phases . this argument is valid for small values of @xmath11 , since the linearization of eq .
( [ kura ] ) is possible only if @xmath102 , but as shown in fig .
[ fig2](a)-(c ) we observe the formation of the same perfectly synchronized clusters of symmetric nodes for a wide range of @xmath11 .
however , when @xmath11 becomes larger than a certain value @xmath26 , the assumption @xmath44 does not hold any more and the global synchronized state loses stability . by looking at fig .
[ fig2](d)-(e ) we notice that for @xmath103 , with @xmath104 for the graph @xmath19 , the value of @xmath66 steadily decreases while the dispersion of phases increases , until it reaches the expected value @xmath105 for a system of seven incoherent oscillators ( see fig . [
fig2](d ) and appendix ) .
moreover , for @xmath106 the maximal lyapunov exponent of the system @xmath35 becomes positive and the system enters a chaotic regime ( see fig . [ fig2](e ) ) .
interestingly , the results reported in fig . [ fig3](a)-(d ) confirm that in this regime the coherence of symmetric nodes , measured by the pairwise order parameter @xmath67 , is higher than expected for incoherent oscillators ( refer to appendix for additional details ) .
figure [ fig3](e ) shows that for @xmath33 the system exhibits metastable , partially synchronized states , in which pairs of symmetric nodes alternates intervals of perfect synchronization with intervals of complete incoherence .
we point out that in this regime chimera states could potentially occur @xcite and could even coexist with remote synchronization for @xmath27 .
qualitatively similar results are obtained for different coupling topologies , but the actual value of @xmath26 seems to depend on the structure of the coupling network in a nontrivial way . _
application to the brain. _ as an example , we investigate here the role of symmetry in the human brain by considering anatomical and functional brain connectivity graphs defined on the same set of @xmath107 cortical areas ( see details in appendix ) .
we have first constructed a graph of anatomical brain connectivity as obtained from dw - mri data @xcite , where links represent axonal fibers , and we used this graph as a backbone network to integrate eqs .
( [ kura ] ) . between pairs of nodes as a function of their phase differences @xmath108 according to the simulated kuramoto dynamics .
the black solid curve corresponds to the average value over all the subjects , while the gray area covers the @xmath109 and the @xmath110 percentiles of the distribution .
the dashed horizontal line indicates the threshold for statistical significant correlations ( @xmath111 , corrected for multiple comparisons).,width=288 ] we identified candidate pairs of anatomically symmetric areas by means of agglomerative clustering ,
i.e. , grouping together nodes having close phases at the stationary state ( full dendrogram and details are provided in appendix ) .
then , we considered the graph of functional brain connectivity , in which links represent statistically significant correlations between the bold fmri time - series of cortical areas ( see details in appendix ) .
figure [ fig4 ] illustrates the results for @xmath31 ( we obtained qualitatively similar results in a wide range of @xmath11 ) .
consider nodes 57 and 74 , corresponding respectively to the green and blue areas in panel ( a ) .
not only the two areas are spatially separated , but there is no edge connecting the two corresponding nodes in the anatomical connectivity network .
however , the two nodes are detected as a candidate symmetric pair since at the stationary state of the kuramoto dynamics in eq .
( [ kura ] ) the oscillators associated to these two nodes have very close phases ( see dendrogram in appendix ) . as shown in fig .
[ fig4](b ) , also the bold fmri signals corresponding to nodes 57 and 74 also are strongly synchronized .
we obtain remarkably different results when we consider node 74 and node 76 .
these nodes correspond to two spatially adjacent areas of the brain ( the red and blue regions in fig .
[ fig4](a ) and are directly connected in the anatomical connectivity network .
however , at the stationary state of eq .
( [ kura ] ) the phase difference of the oscillators associated to node 74 and 76 is quite large .
interestingly , in this case the fmri time - series associated to these nodes are much less similar to each other [ see the two bottom trajectories reported in fig . [ fig4](b ) ] .
to quantify this effect , we plot in fig .
[ fig4]c the average functional correlation @xmath112 between the fmri activity of pairs of brain areas as a function of the phase differences @xmath108 between the phases of the corresponding oscillators , obtained from the dynamics of eq .
( [ kura ] ) on the anatomical connectivity network .
the fact that @xmath112 decreases with @xmath108 suggests that structural symmetry plays an important role in determining human brain functions .
in fact , the functional activities of anatomically symmetric areas can be strongly correlated , even if the areas are distant in space .
these results suggest that the study of anatomical symmetries in neural systems might provide meaningful insights about the functional organization of distant neural assemblies during diverse cognitive or pathological states @xcite .
applied to other connectivity networks as a method to spot potential network symmetries , our study could provide new insights on the interplay between structure and dynamics in complex systems .
the authors thank yasser iturria - medina for sharing the dti connectivity data used in the study , and simone severini for useful comments .
m. v. acknowledges financial support from the spanish ministry of science and innovation , juan de la cierva programme ref .
jci-2010 - 07876 .
a. d - g . acknowledges support from the spanish dgicyt grant fis2009 - 13730 , from the generalitat de catalunya 2009sgr00838 .
this work was supported by the eu - lasagne project , contract no.318132 ( strep ) . |
viscosity , diffusivity , and conductivity are transport coefficients which characterize the dynamics of long wavelength and low frequency fluctuations in a medium .
the quantity shear viscosity ( @xmath0 ) per entropy density ( @xmath12 ) has attracted a lot of attention because of the intriguing conjecture that @xmath4 has a minimum bound of @xmath13 for all systems @xcite .
this conjecture is inspired by the anti - de - sitter space / conformal field theory correspondence ( ads / cft ) @xcite which is rooted in string theory . surprisingly , the hot and dense matter produced at rhic @xcite ( for reviews , see e.g. ref .
@xcite ) just above the phase transition temperature ( @xmath14 ) has @xmath15(theory)@xmath16(experiment ) @xcite , a value close to the conjectured bound .
a robust limit of @xmath17 at @xmath18 was recently extracted from a vishnu hybrid model @xcite and a lattice computation of gluon plasma yields @xmath19 at temperature @xmath20 @xcite .
the qcd transport coefficients have also been studied in other temperatures . when @xmath21 @xmath14 , the @xmath0 of a weakly interacting quark gluon plasma is inversely proportional to the scattering rate , @xmath22 @xcite , where @xmath23 is the strong coupling constant .
the bulk viscosity @xmath1 is suppressed by an additional factor of @xmath24 , arising from the response of the trace of the energy momentum tensor @xmath25 to a uniform expansion .
thus , @xmath1 vanishes when the system is conformal or scale invariant . for a gluon plasma ,
the running of the coupling constant breaks the scale invariance .
thus , @xmath26 , @xmath27 @xcite . in the perturbative region , @xmath28 . when @xmath29 @xmath14 , the effective degrees of freedom are pions . in the chiral limit ( @xmath30 and @xmath31 quarks are massless ) , @xmath32 @xcite and @xmath33 @xcite where @xmath34 is the pion decay constant
. a compilation of perturbative qcd calculation of @xmath0 and @xmath1 can be found , e.g. , in ref .
most of these calculations are performed with finite @xmath2 but zero quark chemical potential @xmath3 .
the purpose of this work is to extend the previous perturbative qcd calculation of @xmath0 and @xmath1 to finite @xmath3 at the leading - log ( ll ) order . at this order
, we find @xmath35 and @xmath27 , same as in the limit of @xmath9 , which give parametrically the dominant contribution at asymptotically large ( compared with the @xmath36 scale where qcd becomes non - perturbative ) @xmath2 or @xmath3 .
the vacuum in our calculation has no spontaneous symmetry breaking , thus , it can not be applied to the color superconducting phase in the @xmath37 limit . in the context of finding the minimal @xmath4 and hence the most perfect fluid
, we explore whether @xmath4 can be further reduced by varying @xmath3 and the quark flavor number @xmath7 in the hope that our perturbative calculation can shed light on the non - perturbative region near @xmath14 where qcd is found to be the most perfect matter ever produced in nature .
@xmath0 and @xmath1 can be calculated through the linearized response function of thermal equilibrium states using the kubo formula . in the leading order ( lo ) expansion in the coupling
constant , the computation involves an infinite number of diagrams @xcite .
however , it have been shown that the summation of the lo diagrams in a weakly coupled @xmath38 theory @xcite or in hot qed @xcite is equivalent to solving the linearized boltzmann equation with temperature - dependent particle masses and scattering amplitudes .
this conclusion is expected to hold in perturbative qcd as well .
the boltzmann equation of a quark gluon plasma describes the evolution of the color and spin averaged distribution function @xmath39 for particle @xmath40 : @xmath41 where @xmath39 is a function of space - time @xmath42 and momentum @xmath43 . for the ll calculation
, we only need to consider two particle scattering processes denoted as @xmath44 with the collision terms in the form @xmath45 .
\label{definition of c ab - cd}\ ] ] where @xmath46 and @xmath47 and @xmath48 where @xmath49 is the matrix element squared with all colors and helicities of the initial and final states summed over .
they are tabulated in table [ amp ] in the appendix .
then the collision term for a quark of flavor @xmath40 is @xmath50 where @xmath51 is the quark helicity and color degeneracy factor and the factor @xmath52 is included when the initial state is formed by two identical particles .
similarly , @xmath53 where @xmath54 is the gluon helicity and color degeneracy factor . in equilibrium , the distributions are denoted as @xmath55 and @xmath56 , with @xmath57 where @xmath2 is the temperature , @xmath30 is the fluid four velocity and @xmath3 is the quark chemical potential .
they are all space time dependent .
the thermal masses of gluon and quark or anti - quark for external states ( the asymptotic masses ) can be computed via arnold:2002zm , mrowczynski:2000ed @xmath58 , \label{mg } \\
m_{q}^{2 } & = & m_{\bar{q}}^{2}=2c_{f}g^{2}\int \frac{d^{3}p}{(2\pi ) ^{3}2e_{p}}(2f_{p}^{g}+f_{p}^{q}+f_{p}^{\bar{q } } ) , \label{thermal mass g}\end{aligned}\]]where @xmath59 , @xmath60 , and @xmath61 .
this yields @xmath62where we have set @xmath63 in the integrals on the right hand sides of eqs .
( [ mg ] ) and ( [ thermal mass g ] ) . the difference from non - vanishing masses is of higher order . to extract transport coefficients , it is sufficient to consider infinitesimal perturbations away from equilibrium which have infinite wave lengths . using the chapman - enskog expansion
we linearize the boltzmann equation to the first order in the derivative expansion in @xmath64 .
thus , we only need the thermal equilibrium distributions @xmath56 and @xmath65 on the left - hand - side of the boltzmann equation .
we can also make use of the zeroth - order energy - momentum conservation relation , @xmath66 , to replace the time derivatives @xmath67 and @xmath68 with spatial gradients : @xmath69 \mathbf{\nabla\cdot u } , \label{dt / dt}\ ] ] where @xmath70 is the baryon number density .
we work in the local rest frame of the fluid element where @xmath71 which implies @xmath72 after taking a derivative on @xmath73 . for the right - hand - side of the boltzmann equation ,
we expand the distribution function of particle @xmath40 as a local equilibrium distribution plus a correction @xmath74 , \label{deviation of f}\ ] ] and @xmath75 can be parametrized as @xmath76}\nabla_{[i}u_{j]}\right ] , \label{khi}\ ] ] where @xmath77 and @xmath78 $ ] means the enclosed indices are made symmetric and traceless .
applying eq.([deviation of f ] ) to the right - hand - side of the boltzmann equation and equating it to left - hand - side , we get linear equations for @xmath79 and @xmath80 . for @xmath81
, we obtain @xmath82}=\frac{e_{g}}{f^{g}f^{g}}\frac{1}{n_{g}}\big [ \frac{1}{2}b_{gg\leftrightarrow gg}^{ij}+\sum\limits_{a=1}^{n_{f}}\left ( b_{gq_{a}\leftrightarrow gq_{a}}^{ij}+b_{g\overline{q}_{a}\leftrightarrow g\overline{q}_{a}}^{ij}+b_{q_{a}\bar{q}_{a}\leftrightarrow gg}^{ij}\right ) \big ] , \label{linear equation for b from gluon}\ ] ] where @xmath83 .
\label{definition b ab - cd}\ ] ] with @xmath84}$ ] and we have suppressed the @xmath85 dependence in @xmath86 and @xmath87 when there is no ambiguity . similarly , for @xmath88 , we obtain @xmath89}&=&\frac{e_{q}^{{}}}{f^{q}f^{q}}\frac{1}{n_{q}}\big [ \frac{1}{2}b_{q_{1}q_{1}\leftrightarrow q_{1}q_{1}}^{ij}+b_{q_{1}\bar{q}_{1}\leftrightarrow q_{1}\bar{q}_{1}}^{ij}+\frac{1}{2}b_{gg\leftrightarrow q_{1}\bar{q}_{1}}^{ij}+b_{q_{1}g\leftrightarrow q_{1}g}^{ij } \nonumber \\ & & + \sum\limits_{a=2}^{n_{f}}\left ( b_{q_{1}q_{a}\leftrightarrow q_{1}q_{a}}^{ij}+b_{q_{1}\bar{q}_{a}\leftrightarrow q_{1}\bar{q}_{a } } ^{ij}+b_{q_{a}\bar{q}_{a}\leftrightarrow q_{1}\bar{q}_{1}}^{ij}\right ) \big ] .
\label{linear equation for b from quark}\ ] ] the corresponding equation for @xmath90 can be obtained from the above equation by interchanging @xmath91 .
the linear equations for @xmath92 is relevant to the shear viscosity computation .
they can be written in a compact form @xmath93 with @xmath94}$ ] taken as the source @xmath95 for the shear viscosity .
similarly , the linear equations for @xmath80 for quark and gluon are given by @xmath96 , \label{linear equation for a from gluon}\end{aligned}\ ] ] and @xmath97 , \label{linear equation for a from quark}\end{aligned}\]]where @xmath98 , @xmath99 and @xmath100 .
\label{definition a ab - cd}\ ] ] these equations are relevant for the bulk viscosity computation . they can be written compactly as @xmath101 using the expression of the thermal mass in eq .
( [ mg ] ) , we can evaluate the prefactor of @xmath102 in eq .
( [ linear equation for a from gluon ] ) as @xmath103 \nonumber
\\ & \equiv & \mathbf{p}^{2}+\tilde{m}_{g}^{2 } , \label{p^2+mg^2}\end{aligned}\ ] ] where @xmath104 and the qcd beta function for @xmath105 is @xmath106 similarly we obtain the prefactor of @xmath102 in eq .
( [ linear equation for a from quark ] ) as @xmath107 \nonumber \\ & \equiv & \mathbf{p}^{2}+\tilde{m}_{q}^{2}. \label{p^2+mq^2}\ ] ] as will be shown later , @xmath108 and @xmath109 , thus the source of the bulk viscosity @xmath110 .
this means @xmath111 is , as expected , proportional to the conformal symmetry breaking .
the energy - momentum tensor of the kinetic theory can be written as @xcite ( note the sign difference in metric with @xcite ) @xmath112 , \label{energy momentum tensor}\]]where @xmath40 sums over the gluons and @xmath113 flavors of quarks and anti - quarks .
this equation expresses the total energy momentum tensor of the system as the sum of individual quasiparticles .
there is no @xmath114 term on the right hand side because energy momentum conservation can not be satisfied unless this term vanishes . in principle
, one expects the form of eq .
( energy momentum tensor ) ( and kinetic theory itself ) will be no longer valid at higher orders in the expansion of coupling constant , however , eq.([energy momentum tensor ] ) does reproduce all the thermal dynamical quantities of qcd at @xmath115 correctly @xcite .
expanding @xmath116 to the first order of @xmath117 , we have the fairly simple expression @xmath118 in deriving the above equation one needs to carefully keep track of the implicit distribution function dependence in @xmath119 through the thermal mass ( [ thermal mass g ] ) .
this expression can then be matched to hydrodynamics . in hydrodynamics , @xmath116 depends on @xmath2 , @xmath3 , and @xmath120 . in the local rest frame of the fluid element with @xmath121 , the most general form of @xmath122 at
the order of one space - time derivative is ( assuming parity and time reversal symmetry ) can be decomposed into shear and bulk viscosity terms @xmath123 ( recall that @xmath67 and @xmath68 can be replaced by @xmath124 as shown in eq.([dt / dt ] ) . ) then using eq.([khi ] ) for @xmath125 and comparing eqs .
( [ energy momentum tensor diviation ] ) and ( [ definition of viscosity ] ) , we obtain @xmath126 } , \label{shear}\ ] ] and @xmath127 they can be written schematically as@xmath128 note that @xmath129 , but we will show @xmath130 later .
we also need the total quark number density @xmath131expanding @xmath132 to the first order of @xmath133 , we have @xmath134eqs .
( [ energy momentum tensor diviation ] , [ delta quark number ] ) will be used in computation of bulk viscosity .
it is well known that @xmath0 should not be negative such that the second law of thermodynamics ( entropy can not decrease in time ) is satisfied .
this requirement is fulfilled by rewriting eqs .
( [ linear equation for b],[vis ] ) as @xmath135 and showing that @xmath0 is quadratic in @xmath136 with a positive prefactor such that it is bounded from below by zero .
indeed , these conditions are satisfied in our expression : @xmath137 where there is no summation over @xmath40 and @xmath138 and we can just take @xmath139 and where @xmath140 ^{2}. \label{definition deta ab - cd}\ ] ] once @xmath0 has the standard quadratic form in @xmath141 , and @xmath142 , we can use the standard algorithm to systematically approach @xmath0 from below @xcite . for bulk viscosity , the collision kernel @xmath143 in eq .
( [ linear equation for a ] ) has two zero modes @xmath144 and @xmath145 which satisfy @xmath146 @xmath144 arises from energy conservation @xmath147 while @xmath145 arises from quark number conservation @xmath148 we can use the landau - lifshitz condition @xmath149and @xmath150to rewrite eq .
( [ bulk ] ) by adding linear combinations of @xmath151 and @xmath152 : @xmath153\equiv \langle a|s_{\zeta } \rangle .
\label{bulk'}\end{aligned}\]]substituting @xmath154 with eq.([linear equation for a ] ) , @xmath1 becomes quadratic in @xmath155 with a positive prefactor:@xmath156 where @xmath157 ^{2}. \label{definition dkhi ab - cd}\ ] ] @xmath158 and @xmath159 in eq.([bulk ] ) can be obtained from thermodynamic relations @xmath160 where the pressure @xmath161 can be read of from eq .
( [ energy momentum tensor ] ) and where @xmath162 , @xmath163 , @xmath164 , @xmath165 and @xmath166 .
we have also used @xmath167 then it can be shown that @xmath168 thus one can solve @xmath169 for @xmath1 using the standard algorithm to systematically approach @xmath1 from below @xcite .
as mentioned in the introduction , the ll result has @xmath170 and @xmath171 .
when @xmath172 , @xmath0 , @xmath1 , and @xmath12 all scale as @xmath173 from dimensional analysis .
for @xmath174 , @xmath0 , @xmath1 , and @xmath12 should be even functions of @xmath3 because they should be invariant under @xmath175 , i.e. the exchange of quarks and antiquarks .
we first show the result for the entropy density @xmath12 . only the leading order @xmath12 for free particles
is needed : @xmath176 where @xmath177 , \\ s_{q,\bar{q}}&=&\int_p\left[\frac{\beta ( e_{p}\mp\mu ) } { e^{\beta(e_{p}\mp\mu)}+1 } + \ln\left ( 1+e^{-\beta ( e_{p}\mp\mu)}\right)\right ] .
\label{sg}\ ] ] in fig .
[ entropy density ] , @xmath178 is shown as a function of @xmath179 .
when @xmath172 , @xmath12 scales as @xmath173 and when @xmath174 , @xmath12 scales as @xmath180 .
this agrees with the expectation that @xmath181 when @xmath182 , and @xmath12 increases with @xmath2 for fixed @xmath3 .
the entropy density @xmath12 also increases monotonically with the number of flavors @xmath7 .
we stop at @xmath10 , just before the asymptotic freedom of qcd is lost when @xmath183 . as functions of @xmath184 for different @xmath7.,width=396,height=250 ] as functions of @xmath184 for different @xmath7 .
, width=397,height=262 ] in fig .
[ eta with musquare ] , we show the normalized shear viscosity @xmath185 as functions of @xmath186 for different @xmath7 .
when @xmath187 , @xmath0 scales as @xmath173 from dimensional analysis . when @xmath188 , @xmath0 can not scale as @xmath189 because it is an even function of @xmath3 . instead
, @xmath0 scales as @xmath190 .
technically , this is because @xmath191 as @xmath188 while the anti - quark and gluon contributions vanish .
thus , eqs .
( [ shear ] ) and ( [ shear ] ) are solved with @xmath192 and @xmath193 physically , this @xmath194 behavior emerges because @xmath0 is inversely proportional to the collision rate which vanishes at @xmath182 . also , @xmath0 is monotonically increasing with @xmath7 because the averaged coupling between gluons is stronger than those with quarks involved .
thus , the effective collision rate is smaller with higher @xmath3 and higher @xmath7 . around @xmath9
, we make a taylor expansion @xmath195 , where @xmath196 is @xmath197 at zero quark chemical potential .
the values of @xmath196 and @xmath198 for various @xmath7 is tabulated in table [ power expansion coefficients eta ] .
our @xmath196 is identical to amy s to at least the second decimal place for all @xmath7 computed in ref .
the agreement is better than @xmath199 . in fig .
[ eta over s with musquare ] , @xmath200 is shown as functions of @xmath186 . for a given coupling , the ll value of @xmath4 is the smallest at @xmath201 , i.e. the fluid is the most perfect at @xmath9 .
this perturbative qcd result at high @xmath2 is consistent with the observation of ref .
@xcite in the hadronic phase at low @xmath2 .
thus , we speculate that this property might also be true near the phase transition temperature @xmath202 such that qcd has its local minimum , perhaps its absolute minimum as well , at @xmath202 with zero quark chemical potential .
if the coupling @xmath203 is held fixed , our @xmath4 is monotonically decreasing with @xmath113 for @xmath8 but not for @xmath11 and @xmath204 ( there is a crossing between the @xmath4 of @xmath205 and @xmath206 at @xmath207 ) . this pattern looks random , but interestingly , it is qualitatively consistent with the pion gas result of ref .
@xcite which has @xmath208 ( we have used @xmath209 ) .
again , this suggests that there is a natural connection between @xmath4 above and below the phase transition .
this pion gas analogy can also explain why @xmath11 and @xmath204 are special
there is no pion in these two cases .
we also observe that when @xmath210 , the fluid can be more perfect than that of @xmath11 . it would be interesting to test this in lattice qcd to find a more perfect fluid than the currently evaluated @xmath11 case
. it would also be interesting to investigate how the above qualitative @xmath7 scaling changes due to the possible infrared fixed point for @xmath211 where chiral symmetry is not supposed to be broken and hence no pions exist anymore ( see , e.g. @xcite and references therein ) . as functions of @xmath186 for different @xmath7.,width=398,height=263 ] as functions of @xmath186 for different @xmath7.,width=400,height=270 ] as functions of @xmath186 for different @xmath7.,width=404,height=269 ] in fig .
[ bulk viscosity with musquare ] we show the normalized bulk viscosity @xmath212 as functions of @xmath186 for different @xmath7 .
when @xmath187 , @xmath1 scales as @xmath173 from dimensional analysis .
when @xmath188 , @xmath1 scales as @xmath180 .
technically , this is because as @xmath213 , @xmath214 , the dominant contribution in eq.([bulk ] ) comes from the scattering between quarks . in quark scattering , the combination @xmath215 ^{2}=\mathcal{o}(t^{2})$ ] in eq.([definition dkhi ab - cd ] ) .
this is because when @xmath213 , the scattering can only happen on the fermi surface otherwise it will be pauli blocked ( this is imposed by the vanishing of the prefactor @xmath216 ) . but scattering on the fermi surface yields @xmath217 and thus @xmath218
. therefore , the dimensionless combination @xmath219 contributes at @xmath220 in eq.([definition dkhi ab - cd ] ) , which leads to @xmath221 .
( a similar argument can be applied to @xmath222 . as @xmath213 , @xmath223 ^{2}=\mathcal{o}(t^{0})$ ] in eq.([definition deta ab - cd ] ) , since the scattering on the fermi surface does not have to be forward scattering thus @xmath224 does not have to vanish .
this leads to @xmath225 in eq.([definition deta ab - cd ] ) . )
in contrast to @xmath0 , @xmath1 is not monotonically increasing with @xmath7 because @xmath1 is suppressed by an additional power of @xmath226 .
thus when @xmath7 is small , @xmath1 increases with @xmath7 because quarks tend to make the averaged effective coupling weaker but at large @xmath7 , the suppression factor @xmath227 takes control to make @xmath1 decrease with @xmath7 .
the maximum @xmath1 happens when @xmath228 or @xmath229 , depending on the value of @xmath230 . around @xmath9
, we make a taylor expansion @xmath231 , where @xmath232 is @xmath233 at zero quark chemical potential .
the values of @xmath232 and @xmath234 for varies @xmath7 is tabulated in table [ power expansion coefficients bulk ] .
our @xmath232 is identical to adm s at least to the second decimal place for all @xmath7 computed in ref .
the agreement is better than @xmath199 . in fig .
[ bulk over s with musquare ] , if the coupling @xmath203 is fixed , our @xmath6 is monotonically decreasing with @xmath7 .
this pattern is qualitatively consistent with the massless pion gas result of ref .
@xcite which has @xmath235 ( only valid for @xmath236 where pions exist ) .
we have calculated the shear and bulk viscosities of a weakly coupled quark gluon plasma at the leading - log order with finite temperature @xmath2 and quark chemical potential @xmath3 .
we have found that when normalized by the entropy density @xmath12 , @xmath4 increases monotonically with @xmath3 and eventually scales as @xmath5 at large @xmath3 .
however @xmath6 is insensitive to @xmath3 .
both @xmath4 and @xmath6 are monotonically decreasing function of the quark flavor number @xmath7 when @xmath236 .
the same property is also observed in pion gas calculations .
our perturbative calculation suggests that qcd becomes the most perfect ( with the smallest @xmath4 ) at @xmath9 and @xmath10 ( the maximum @xmath7 with asymptotic freedom ) .
it would be interesting to test whether the currently smallest @xmath4 computed close to the phase transition at @xmath9 and @xmath11 can be further reduced by increasing @xmath7 .
acknowledgement : jwc and yfl thank the int , seattle , for hospitality .
jwc and yfl are supported by the nsc , ncts , and casts of roc .
yks is supported in part by the ccnu - qlpl innovation fund ( qlpl2011p01 ) .
qw is supported in part by the national natural science foundation of china under grant no .
[ c]|l|l|@xmath237 & @xmath238 + @xmath239{c } q_{1 } q_{2}\rightarrow q_{1 } q_{2}\\ \bar{q}_{1 } q_{2}\rightarrow\bar{q}_{1 } q_{2}\\ q_{1 } \bar{q}_{2}\rightarrow q_{1 } \bar{q}_{2}\\ \bar{q}_{1 } \bar{q}_{2}\rightarrow\bar{q}_{1 } \bar{q}_{2}\end{array } $ ] & @xmath240 + @xmath239{c } q_{1 } q_{1}\rightarrow q_{1}q_{1}\\ \bar{q}_{1 } \bar{q}_{1}\rightarrow\bar{q}_{1}\bar{q}_{1}\end{array } $ ] & @xmath241 + @xmath242 & @xmath243 + @xmath244 & @xmath245 + @xmath246 & @xmath247 + @xmath239{c } q_{1 } g\rightarrow q_{1}g\\ \bar{q}_{1 } g\rightarrow\bar{q}_{1}g \end{array } $ ] & @xmath248 + @xmath249 & @xmath250 + [ c]|l|l|l||l|l|l|@xmath7 & @xmath196 & @xmath198 & @xmath7 & @xmath196 & @xmath198 + 0 & 27.125 & 0 & 9 & 172.564 & 50.381 + 1 & 60.808 & 16.619 & 10 & 178.839 & 51.301 + 2 & 86.472 & 27.281 & 11 & 184.389 & 52.028 + 3 & 106.664 & 34.454 & 12 & 189.333 & 52.608 + 4 & 122.957 & 39.459 & 13 & 193.764 & 53.074
+ 5 & 136.380 & 43.055 & 14 & 197.760 & 53.450 + 6 & 147.627 & 45.703 & 15 & 201.380 & 53.755 + 7 & 157.187 & 47.690 & 16 & 204.675 & 54.003 + 8 & 165.412 & 49.207 & & & + [ c]|l|l|l||l|l|l|@xmath7 & @xmath232 & @xmath234 & @xmath7 & @xmath232 & @xmath234 + 0 & 0.4430 & 0 & 9 & 0.3847 & 0.0873 + 1 & 0.5816 & 0.0393 & 10 & 0.3113 & 0.0725 + 2 & 0.6379 & 0.0747 & 11 & 0.2389 & 0.0569 + 3 & 0.6568 & 0.0981 & 12 & 0.1706 & 0.0414 + 4 & 0.6495 & 0.1116 & 13 & 0.1096 & 0.0270 + 5 & 0.6218 & 0.1172 & 14 & 0.0592 & 0.0148 + 6 & 0.5778 & 0.1163 & 15 & 0.0225 & 0.0057 + 7 & 0.5213 & 0.1103 & 16 & 0.0026 & 0.0007 + 8 & 0.4558 & 0.1003 & & & + to describe the microscope scattering processes in a quark gluon plasma , we need scattering amplitudes between quarks and gluons . in a hot qcd plasma the infrared singularity in the amplitude can be regularized by hard thermal loop dressed propagator .
we use the same amplitude ( shown in table [ amp ] ) as ref .
@xcite .
around @xmath9 , we make taylor expansions @xmath251 and @xmath252 . the values of @xmath196 , @xmath198 , @xmath232 and @xmath234 for various @xmath7 are tabulated in table [ power expansion coefficients eta ] and [ power expansion coefficients bulk ] .
our @xmath196 and @xmath232 are identical to amy s and adm s to at least the second decimal place for all @xmath7 computed in ref .
@xcite and @xcite , respectively .
the agreement is better than @xmath199 .
k. adcox _ et al . _
[ phenix collaboration ] , nucl .
phys . a * 757 * , 184 ( 2005 ) [ nucl - ex/0410003 ] . b. b. back , m. d. baker , m. ballintijn , d. s. barton , b. becker , r. r. betts , a. a. bickley and r. bindel _ et al .
_ , nucl .
phys . a * 757 * , 28 ( 2005 ) [ nucl - ex/0410022 ] .
j. adams _ et al . _ [ star collaboration ] , nucl . phys . a * 757 * , 102 ( 2005 ) [ nucl - ex/0501009 ] .
j. -w .
chen , j. deng , h. dong and q. wang , arxiv:1107.0522 [ hep - ph ] .
j. -w .
chen , j. deng , h. dong and q. wang , phys .
d * 83 * , 034031 ( 2011 ) [ erratum - ibid .
d * 84 * , 039902 ( 2011 ) ] [ arxiv:1011.4123 [ hep - ph ] ] .
s. mrowczynski and m. h. thoma , phys .
d * 62 * , 036011 ( 2000 ) [ hep - ph/0001164 ]
. j. i. kapusta , nucl . phys .
b * 148 * , 461 ( 1979 ) .
j. -w .
chen , y. -h .
li , y. -f .
liu and e. nakano , phys . rev .
d * 76 * , 114011 ( 2007 ) [ hep - ph/0703230 ] . | we calculate the shear ( @xmath0 ) and bulk ( @xmath1 ) viscosities of a weakly coupled quark gluon plasma at the leading - log order with finite temperature @xmath2 and quark chemical potential @xmath3 .
we find that the shear viscosity to entropy density ratio @xmath4 increases monotonically with @xmath3 and eventually scales as @xmath5 at large @xmath3 . in contrary
, @xmath6 is insensitive to @xmath3 .
both @xmath4 and @xmath6 are monotonically decreasing functions of the quark flavor number @xmath7 when @xmath8 .
this property is also observed in pion gas systems .
our perturbative calculation suggests that qcd becomes the most perfect ( i.e. with the smallest @xmath4 ) at @xmath9 and @xmath10 ( the maximum @xmath7 with asymptotic freedom ) .
it would be interesting to test whether the currently smallest @xmath4 computed close to the phase transition with @xmath9 and @xmath11 can be further reduced by increasing @xmath7 . |
gamma - ray bursts ( grbs ) are among the most powerful explosions in the universe .
it is generally believed that the impulsively injected fireball results from core collapse in a massive star ( woosley 1993 ; macfadyen & woosley 1999 ) , or from the merging of either two neutron stars or a neutron star and a black hole ( e.g. , ruffert et al . 1997 ; popham et al . 1999 ; narayan et al .
1992 , 2001 ; rosswog & davies 2002 ; lee & ramirez - ruiz 2002 ) .
after the explosion , the relativistic ejecta collide with the ambient interstellar medium causing x - ray , optical , and radio emission .
these so - called `` afterglows '' thus carry important information on the injection mechanism , the configuration of the ( possibly collimated ) fireball , and the surrounding environment ( e.g. , mszros 2002 ) .
two kinds of grbs have been defined according to whether their gamma - ray emission has duration longer or shorter than 2 s. although their frequency distributions overlap , that of the short - duration grbs peaks at 0.3 s , while that of the long - duration grbs peaks at 3040 s ( kouveliotou et al .
in addition , the duration is weakly correlated with the spectral hardness ratio at high energies : short grbs tend to be harder and long grbs tend to be softer ( kouveliotou et al .
1993 ) .
the optical afterglows of short grbs were elusive until the detection ( fenimore et al .
2004 ) of grb 040924 by the high energy transient explorer 2 ( _ hete2 _ ) on 2004 sep .
24 , at 11:52:11 ( ut dates are used throughout this paper ) .
this event lasted about 1.2 s and was x - ray rich according to the _
hete2 _ flux in the 730 kev and 30400 kev bands .
the konus - wind satellite also detected this event with 1.5 s duration in the 20300 kev band ( golenetskii et al .
2004 ) . since the high - energy spectrum of grb 040924 is soft ( fenimore et al .
2004 ) , the object might actually be near the short - duration end of the long grbs .
a detailed study of the associated optical afterglow could provide further information on whether this is indeed the case , thus probing the nature of grbs in the boundary region . about @xmath4 min after the burst , fox ( 2004 ) detected the corresponding optical afterglow with an @xmath0-band magnitude of @xmath518 .
this was shortly followed by li et al .
( 2004 ) , who reported @xmath6 mag at 26 min after the burst . from then on , a number of observatories joined in the follow - up observations ( fynbo et al .
2004 ; hu et al . 2004
; khamitov et al .
2004a ; terada , akiyama , & kawai 2004 ) .
radio observations failed to detect the afterglow at @xmath7 hr and @xmath8 d ( frail & soderberg 2004 ; van de horst , rol , & wijers 2004a , b ) .
spectral measurements by the very large telescope ( vlt ) of a galaxy located at the position of the optical afterglow indicated a redshift @xmath9 for this event ( wiersema et al .
upon receiving the burst alert message from _
hete2 _ and the optical position reported by fox & moon ( 2004 ) , multi - band ( johnson @xmath10 and @xmath11 ;
bessell @xmath0 and @xmath12 ) follow - up observations of grb 040924 with the lulin one - meter telescope ( lot , in taiwan ) were initiated according to the previously approved target - of - opportunity procedure .
photometric images were obtained with the pi1300b ccd camera ( @xmath13 pixels , @xmath14 field of view ; kinoshita et al .
2005 ) during the interval 14.3120.89 on sep . 24 ( i.e. , 2.49.0 hr after the burst ) .
unfortunately , the earliest observations ( @xmath15 hr ) were not successful because of poor weather and short exposure times .
these problems also affected all of the @xmath10 and @xmath12 data , and many of the @xmath11 and @xmath0 images as well .
nevertheless , our observations reveal unusual early - time evolution of the afterglow brightness , as discussed below .
a standard routine including bias subtraction and flat - fielding corrections with appropriate calibration data was employed to process the data using iraf .
the afterglow was clearly seen in the @xmath11-band and @xmath0-band images ( figure 1 ) .
the position of the afterglow is @xmath16(j2000 ) @xmath17 , @xmath18(j2000 ) @xmath19 ( @xmath20 in each coordinate ) .
next , the daophot package ( stetson 1989 ) was used to perform aperture photometry of the grb field by choosing ten field stars for differential photometry .
the lot data were combined with median filtering to improve the signal - to - noise ratio . for the photometry ,
the aperture diameter was set to 4 times the fwhm of the objects .
the magnitude error was estimated as @xmath21 , where @xmath22 is the photometric error of the afterglow estimated from the daophot output , and @xmath23 is the systematic calibration error estimated by comparing the instrumental magnitudes of the ten field stars .
besides the calibration data obtained by the usnofs 1.0-m telescope ( henden 2004 ) , we used the measurements of four landolt ( 1992 ) standard - star fields ( sa92 , pg2331 + 055 , sa95 , and pg2317 + 046 ) taken by lot on a photometric night .
the difference between the two flux calibrations is within 0.04 mag .
the magnitudes derived for the @xmath0 and @xmath11 observations are summarized in table 1 .
in addition to the lot data , we have also included two early measurements from the 0.76-m katzman automatic imaging telescope ( kait ; li et al .
2003b ) at lick observatory at @xmath24 and 1.06 hr .
the kait data were taken without filters , but the transformation of the unfiltered magnitude to @xmath0 can be determined from the @xmath25 colors of the grb field stars and of the optical afterglow ( li et al .
2003a , b ) .
the calibration of the grb 040924 field is adopted from henden ( 2004 ) , and the value @xmath26 mag of the afterglow is taken from lot data at 0.292 d after the burst .
kait observed the grb at low airmass ( 1.261.4 ) , and the local standard stars have @xmath25 colors ( 0.390.85 mag ) similar to that of the grb .
moreover , from three photometric nights we found that the coefficient for the second - order extinction is only 0.04 ; thus , the errors caused by second - order extinction of grb 040924 are small , and are included in the overall uncertainties of the kait data .
the light curve of grb 040924 in figure 2 is a combination of the early observations reported by fox ( 2004 ) at 0.012 d and 0.033 d after the burst , the work reported here , and the measurements by khamitov et al .
( 2004b , c ) at @xmath27 d , 0.62 d , and 1.56 d , fynbo et al .
( 2004 ) at @xmath28 d , and silvey et al .
( 2004 ) around @xmath29 d. to put all of the data onto a consistent magnitude scale , we recalibrated the above - mentioned published data by using the henden ( 2004 ) standard stars for the grb 040924 field .
the data of fox ( 2004 ) were calibrated by gsc 2.2 stars with f - emulsion magnitude which corresponds closely to the @xmath0-band magnitude ; the gsc stars are 0.11 mag brighter than the henden standard stars in the average of our images . since two reference stars are provided by khamitov et al .
( 2004b , c ) and fynbo et al .
( 2004 ) , we measured these stars from lot @xmath0-band combined images and obtained the average magnitudes and root - mean - square errors ; the results were then used to recalibrate the reported afterglow magnitudes .
the time evolution of the light curve can be expressed in terms of a power law with @xmath30 , where @xmath31 is the time after the burst and @xmath16 is the index .
we find @xmath32 ( @xmath33= 0.06 for @xmath34 ) for the very sparse @xmath11-band data ( only three closely spaced lot observations and one later observation from silvey et al .
similarly , we derive @xmath35 ( @xmath33 = 2.08 for @xmath36 ) from all 14 available @xmath0-band observations .
these two values of @xmath16 fall within the range of long grbs ( @xmath37 to @xmath38 ) , so the afterglow of grb 040924 is consistent with the standard model of cosmic - ray electrons accelerated by the internal and external shocks of the expanding fireball @xcite , as in the case of typical long - duration grbs . upon closer scrutiny ,
the first three @xmath0-band observations ( at @xmath450 min ) indicate @xmath1 , consistent with the conclusion of fox ( 2004 ) , while the subsequent data ( starting from the third @xmath0-band observation ) give a somewhat steeper value of @xmath39 ( with @xmath40 for @xmath41 ) .
[ essentially the same late - time result , @xmath42 , is found when we use only our own lot and kait data , together with the sibley et al .
( 2004 ) observation at @xmath43 d. ] the data thus suggest the presence of a mild break , the significance of which is discussed below . finally , our lot observations of grb 040924 at @xmath44 hr indicate a color index of @xmath45 mag , corrected for foreground reddening of @xmath46 mag ( schlegel , finkbeiner , & davis 1998 ) . we have also calculated the color of observations by silvey et al .
( 2004 ) at @xmath43 d ( 22.1 hr ) to be @xmath47 mag , corrected for foreground reddening . while these two values are consistent with the color of typical long grbs ( @xmath48 mag ; simon et al .
2001 ) , they also may suggest the interesting possibility of a color change during the time evolution of this afterglow .
however , the color change is only marginally significant , given the uncertainties .
future grb afterglow measurements should shed new light on this tantalizing behavior .
the brightness variations of the optical afterglows of grbs potentially yield important information on the expansion of the ejecta . for example , breaks in the light - curve power laws at several hours to several days after the bursts have been observed in a number of grbs .
this effect is generally believed to be associated with the evolution as a collimated jet ( rhoads 1999 ) . in the case of grb 040924 , because of the small variation from @xmath49 to @xmath3 around @xmath50 min , the break is not well constrained .
on the other hand , this small break could be indicative of some interesting physical process .
note that the amplitude of the break ( @xmath51 ; here @xmath52 and @xmath53 are the power - law indices before and after the break , respectively ) is independent of extinction under the assumption of no color change . in the case of grb 040924 , from @xmath54 and @xmath55
we find @xmath56 .
according to theoretical work ( rhoads 1999 ) , @xmath57 for a collimated jet with a fixed angle , and @xmath58 for a sideways - expanding jet in the framework of a constant ambient density model .
it is clear that the amplitude of the break in grb 040924 is much smaller than values expected of jet expansion with power - law indices much steeper after the break .
the interpretation of a jet break for grb 040924 is thus uncertain .
next we will explore an alternative possible explanation .
panaitescu & kumar ( 2000 ) pointed out that a light - curve break could also be caused by the spectral cooling frequency moving through the optical band .
this property might be used as a diagnostic tool to differentiate among different possible scenarios of grb afterglow formation . in the standard grb afterglow model
, it is usually assumed that the synchrotron emission observed in optical bands originates from the expansion of a blast wave of constant energy into an interstellar medium ( ism ) of constant density .
however , there is also increasing evidence that some grbs have massive - star progenitors .
consequently , the corresponding relativistic blast waves should actually be expanding into the stellar wind of the progenitor stars with a density variation of @xmath59 ( dai & lu 1998 ; mszros et al .
1998 ; panaitescu et al .
zhang & mszros ( 2004 ) listed the broad - band optical spectra of the synchrotron radiation from a power - law distribution of energetic electrons with a spectral index ( @xmath60 ) accelerated by the blast wave ; accordingly , we can obtain the values of @xmath60 before and after the cooling break .
as shown in table 2 , the ism model provides the only possible fit ( for @xmath61 ) to the grb 040924 observations which satisfies the requirement that @xmath60 should remain nearly the same ( @xmath62 to 2.08 ) as the spectrum evolves from @xmath63 to @xmath64 .
note that within the framework of the ism model , @xmath65 for @xmath63 and @xmath66 for @xmath64 if the corresponding light curve can be characterized by a single power - law index ( @xmath67 ) .
another interesting estimate can be made concerning the relation between the cooling - break frequency @xmath68 , the break time @xmath69 ( in units of days ) , the redshift @xmath70 , the magnetic energy @xmath71 , the kinetic energy @xmath72 ( in units of @xmath73 erg ) , and the density @xmath74 of the ism .
according to granot & sari ( 2004 ) , @xmath75 now , with @xmath76 , @xmath77 hz in the @xmath0 band , @xmath9 , @xmath78 , and the assumptions that @xmath79 and @xmath80 @xmath81 , we find @xmath82 .
this value is consistent with the normal assumption for the magnetic - energy fraction of @xmath83 , though slightly larger . for the case of a single power - law index ( @xmath35 ) , @xmath84 , which is much smaller than the normal value .
our analysis thus suggests that the observed light curve of grb 040924 could be the result of the spectral cooling frequency moving through the optical band . in other words , the apparent break in grb
040924 might not be a jet break but rather a cooling break .
the 1.21.5 s duration of grb 040924 , though formally in the domain of short grbs , overlaps the short end of long - duration grbs .
moreover , it has a soft high - energy spectrum , characteristic of long grbs .
our optical afterglow observations show that the temporal evolution , power - law index , and @xmath25 color of grb 040924 are also consistent with those of well - observed long grbs .
the signature of a low - amplitude break in the light curve , as suggested by our present data , can be explained by the afterglow model invoking a cooling break at early times .
however , note that the jet break usually occurs 12 d after the burst , and there are few observations of grb 040924 at @xmath85 d. thus , we can not exclude the possibility that the true jet break occurred outside the range of our observations . due to the general lack of information on the optical afterglows of short grbs , we can not compare our observations of grb 040924 to this class of objects .
the @xmath86 satellite , with higher gamma - ray sensitivity and more accurate localization than previous missions , will provide more opportunities to understand the properties of typical short grbs and of grbs near the boundary between short and long grbs .
we thank the staff and observers at the lulin telescope for various arrangements that made possible the observations reported herein .
this work is supported by grants nsc 93 - 2752-m-008 - 001-pae and nsc 93 - 2112-m-008 - 006 .
y.u . acknowledges support from the japan society for the promotion of science ( jsps ) through a jsps research fellowship for young scientists .
is grateful for nsf grant ast-0307894 , and for a miller research professorship at uc berkeley during which part of this work was completed .
kait was made possible by generous donations from sun microsystems , inc . , the hewlett - packard company , autoscope corporation , lick observatory , the nsf , the university of california , and the sylvia and jim katzman foundation .
dai , z. g. , & lu , t. 1998 , , 298 , 87 fenimore , e. e. , et al .
2004 , gcn circ .
2735 fox , d. b. 2004 , gcn circ .
2741 fox , d. b. , & moon , d. s. 2004 , gcn circ .
2734 frail , d. a. , & soderberg , a. 2004 , gcn circ .
2758 fynbo , j. p. u. , hornstrup , a. , hjorth , j. , jensen , b. l. , & anderson , m. i. 2004 , gcn circ .
2747 henden , a. 2004 , gcn circ .
2811 hu , j. h. , lin , h. c. , huang , k. y. , urata , y. , ip , w. h. , & tamagawa , t. 2004 , gcn circ .
2744 golenetskii , s. , aptekar , r. , mazets , e. , palshin , v. , frederiks , d. , & cline , t. 2004 , gcn circ .
2754 khamitov , i. , et al . 2004a , gcn circ .
2740 khamitov , i. , et al . 2004b , gcn circ .
2749 khamitov , i. , et al . 2004c , gcn circ .
2752 kinoshita , d. , et al .
2005 , chjaa , 5 , 315 kouveliotou , c. , et al .
1993 , , 463 , 570 lee , w. h. , & ramirez - ruiz , e. 2002 , , 577 , 893 li , w. , filippenko , a. v. , chornock , r. , & jha , s. 2003a , apj , 586 , l9 li , w. , filippenko , a. v. , chornock , r. , & jha , s. 2003b , pasp , 115 , 844 li , w. , filippenko , a. v. , chornock , r. , & jha , s. 2004 , gcn circ .
2748 landolt , a. u. 1992 , , 104 , 340 macfadyen , a. i. , & woosley , s. e. 1999 , , 524 , 262 mszros , p. , rees , m. j. , & wijers , r. a. m. j. 1998 , , 499 , 301 mszros , p. 2002
, , 40 , 137 narayan , r. , et al .
1992 , , 395 , 83 narayan , r. , et al .
2001 , , 557 , 494 panaitescu , a. , & kumar , p. 2000 , , 543 , 66 panaitescu , a. , mszros , p. , & rees , m. j. 1998 , , 503 , 314 popham , r. 1999 , , 518 , 356 rhoads , j. e. 1999 , , 525 , 737 rosswog , s. , & davies , m. b. 2002 , , 334 , 481 ruffert , m. , et al . 1997 , a&a , 319 , 122 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , apj , 500 , 525 silvey , j. , et al .
2004 , gcn circ .
2833 simon , v. 2001 , a&a , 377 , 450 stetson , p. b. 1987 , pasp , 99 , 191 terada , h. , akiyama , m. , & kawai , n. 2004 , gcn circ .
2745 van der horst , a. j. , rol , e. , & wijers , r. a. m. j. 2004a , gcn circ .
2746 van der horst , a. j. , rol , e. , & wijers , r. a. m. j. 2004b , gcn circ .
2759 wiersema , k. , et al .
2004 , gcn circ .
2800 woosley , s. e. 1993 , , 405 , 273 zhang , b. , & mszros , p. 2004 , int . j. mod .
a. , 19 , 2385 ccccccc ut date & start time & mean delay ( days ) & filter & exposure ( s ) & mag & site + 2004 - 09 - 24 & 18:52:37 & 0.296 & @xmath11 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 19:46:02 & 0.333 & @xmath11 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 20:24:54 & 0.360 & @xmath11 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 12:18:21 & 0.018 & @xmath0 & 120 s @xmath87 1 & [email protected] & kait + 2004 - 09 - 24 & 12:55:21 & 0.044 & @xmath0 & 120 s @xmath87 1 & [email protected] & kait + 2004 - 09 - 24 & 15:00:55 & 0.140 & @xmath0 & 600 s @xmath87 2 & [email protected] & lot + 2004 - 09 - 24 & 18:34:37 & 0.284 & @xmath0 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 19:28:57 & 0.321 & @xmath0 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 20:07:48 & 0.348 & @xmath0 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 24 & 20:42:17 & 0.372 & @xmath0 & 300 s @xmath87 3 & [email protected] & lot + 2004 - 09 - 25 & 08:35:00 & 0.873 & @xmath0 & 300 s @xmath87 3 & @xmath8922.47 & kait + ccccccccc & & & & & & & & + & & & & & & & & + @xmath90 & ism & @xmath91 & 1.93 & 2.41 & & @xmath92 & @xmath93 & @xmath94 + @xmath95 & ism & @xmath96 & 1.60 & 2.08 & & @xmath97 & 1.73 & 3.65 + @xmath98 & wind & @xmath99 & 1.26 & 1.74 & & @xmath100 & @xmath101 & 2.48 + @xmath95 & wind & @xmath96 & 1.6 & 2.08 & & @xmath102 & @xmath103 & 0.48 + @xmath98 & jet & @xmath104 & 0.7 & 1.06 & & @xmath105 & @xmath106 & @xmath107 + @xmath95 & jet & @xmath104 & 0.7 & 1.06 & & @xmath105 & @xmath106 & @xmath107 + | the 1-m telescope at lulin observatory and the 0.76-m katzman automatic imaging telescope at lick observatory were used to observe the optical afterglow of the short - duration ( 1.21.5 s ) gamma - ray burst ( grb ) 040924 .
this object has a soft high - energy spectrum , thus making it an exceptional case , perhaps actually belonging to the short - duration tail of the long - duration grbs .
our data , combined with other reported measurements , show that the early @xmath0-band light curve can be described by two power laws with index @xmath1 ( at @xmath250 min ) and @xmath3 ( at later times ) .
the rather small difference in the spectral indices can be more easily explained by an afterglow model invoking a cooling break rather than a jet break . |
close binaries of compact solar - mass objects are expected to form via the evolution of massive star binaries or by dynamical interaction in dense star clusters .
neutron star ( ns ns ) binaries have been detected as radio pulsars ( e.g. * ? ? ?
* ) , and while black hole neutron star ( bh - ns ) or double black hole ( bh - bh ) binaries have not been observed directly , they are predicted by population synthesis models .
the compact objects are expected to merge due to gravitational wave emission , with evolutionary scenarios estimating a local rate of ns
ns mergers @xmath4 times higher than for bh - ns and bh - bh systems ( e.g. * ? ? ?
* ) . the final remnant for ns ns and ns bh coalescence is generally thought to be a bh of a few solar masses surrounded by a @xmath5 accreting disc ( e.g. * ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the accretion power immediately following the merger is perhaps the ultimate cause of sgrbs @xcite . at early times ( @xmath6 sec )
, the accreting disc is geometrically thin , effectively cooled by neutrino emission @xcite . when the accretion rate drops below @xmath7 the exact value depending on the accretion parameter @xmath8 and bh spin ( chen & beloborodov 2007 ; metzger , piro & quataert 2008 ) the disc becomes radiatively inefficient and super - eddington accretion drives a substantial outflow @xcite . during the dynamical phase of the merger , in which the lighter companion is tidally disrupted , a fraction ( @xmath9 ) of the debris receives enough energy to be ejected from the system while a comparable amount remains bound in eccentric orbits ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and will eventually return to the disc site : _ fallback _ matter .
this weakly bound matter may give rise to interesting phenomena observable on timescales longer than any viscous timescale of the disc .
for example , it has been suggested @xcite that it can be responsible for the x - ray flaring , observed in sgrb afterglows on timescales of minutes to hours ( e.g. * ? ? ? * ) .
unfortunately , numerical investigations have not yet been able to follow the long - term ( @xmath10 minutes ) evolution of this eccentric tail , because of time - step limitations @xcite . in this _ letter _ , we investigate analytically the fate of matter falling back onto a recent merger .
we argue that energy released during fallback is _ not _ a promising source of the x - ray flares . the energy liberated during fallback will either lead to a powerful , radiation - driven wind or a more gradually expanding `` breeze '' that could ultimately form a bound cloud around the merged object . in either case , the expanding gas is so opaque that the radiation is trapped in the expanding flow and degraded to low energies before being released in the x - ray band .
we therefore suggest that compact binary mergers might be accompanied by delayed x - ray emission .
we assess the detectability of this emission when the merger is localized by either a short @xmath1-ray burst or a gravitational wave signal .
a direct observation of the accretion activity would give us valuable information on how compact - object binaries merge .
this _ letter _ is organized as follows .
we discuss the behavior of the fallback matter in [ sec : fb ] .
then , we consider two possible scenarios for this material as it rebounds : we model a wind in
[ sec : wind ] and a bound atmosphere in [ sec : atmo ] .
prospects for detecting the x - ray emission are discussed in [ sec : detection ] and conclusions are drawn in [ sec : conclu ] .
in our analysis , we scale our parameters with values appropiate for ns - ns binaries , since these systems are the most common .
the encounter of a couple of neutron stars is followed by the formation of a central attractor with typical mass @xmath11 @xcite , surrounded by an accretion disc that extends initially up to @xmath12 ( e.g. , * ? ? ?
the weakly bound material , @xmath13 @xcite , launched into elliptical orbits , will travel as far as its apocenter and eventually come back to its pericenter @xmath14 .
the rate at which this material accretes can be found analytically assuming that the energy distribution with mass is flat : the accretion rate , after a plateau phase at @xmath15 , decreases with time as _
fb(t ) = _ max ( ) ^5/3 , [ eq : mdotfb ] @xcite , where the minimum arrival time corresponds to the period of orbits with eccentricity @xmath16 , t_min 10 ^ -2 m_c^-1/2 r_d_7 ^ 3/2 sec , and the initial accretion occurs at a rate _ max = 2 .
[ eq : dotmax ] even if eq .
[ eq : mdotfb ] has been originally derived for tidal distruption of stars in the potential well of a supermassive black hole , numerical calculations by @xcite indicate that the @xmath17 law also applies to the case we are investigating .
when the fallback matter hits the disc ( or the leftover material ) , its kinetic energy ( per unit mass ) @xmath18 is converted into heat via shocks .
the internal energy of the shocked matter is photon - dominated .
initially , the fallback matter would simply join the disc and accrete onto the central object , because it is effectively cooled by neutrino emission : i.e. , the flow rate is _
sub_-eddington with respect to the neutrino luminosity and accretion is possible .
the neutrino emissivity @xmath19 is due both to electron - positron pair annihilation @xmath20 and capture onto nuclei @xmath21 ( see * ? ? ?
* for the analytic approximations ) .
the fallback matter density at @xmath22 is @xmath23 , while the temperature to which the gas is shock - heated can be approximately obtained by equating its kinetic energy density at @xmath22 , ( @xmath24 ) , to its internal energy density , t_sh = ( ) ^1/4 = 3.6 10 ^ 10 ( ) ^5/12k , [ eq : tsh ] where @xmath25 is the radiation constant .
the bh feeding happens for large enough accretion rates , @xmath26 , when the cooling time @xmath27 is shorter than the viscous time at @xmath22 @xmath28 sec .
when @xmath29 sec , the accretion rate @xmath30 drops below the critical value @xmath31 and neutrino cooling becomes inefficient and eventually ( when @xmath32 ) switches off completely .
the remaining reservoir of mass in the tail is still substantial @xmath33 and , unable to accrete onto the black hole , it is likely to be blown off the disc plane .
the first possible fate for the fallback matter that we consider is the formation of a radiation - driven wind .
the amount of mass entrained in the wind @xmath34 and its specific energy are uncertain .
one possibility is that the total kinetic energy of fallback matter is deposited unevenly , so that a small fraction of mass ( @xmath35 ) can reach a final velocity that exceeds the escape velocity and form a wind . on the other extreme
, the wind could have an amount of mass _ comparable _ to the fallback tail @xmath36 , where sufficient internal energy to unbind this weakly bound matter is gained via accretion .
given the range of uncertainty , we will scale our equations adopting @xmath37 and a terminal velocity @xmath38 , where c is the speed of light and we use as guidance the escape velocity @xmath39 at @xmath40 . to model the wind , we take an initial radius @xmath41 .
this is sufficiently close to the sonic radius that we may assume , in first approximation , an outflow with constant velocity equal to its terminal velocity @xmath42 the velocity at the sonic point is only @xmath43 smaller than the terminal velocity . ] .
the wind , powered by fallback matter , will steadily decrease with time according to eq .
[ eq : mdotfb ] , @xmath44 where @xmath45 sec and the time @xmath46 since the onset of the wind is in hours ( @xmath47 ) .
its matter density then follows from matter conservation , ( r , t ) = , for radii @xmath48 .
the radiation pressure @xmath49 can be related to @xmath50 by the polytropic relation , with index @xmath51 .
therefore , the temperature decreases slowly as t(r , t ) p^1/4 r^-2/3 t^-5/12 .
[ eq : trt ] the radiation transported with the wind is mostly liberated at the trapping radius @xmath52 where the diffusion timescale for photons equals the expansion timescale . beyond this radius
, the luminosity is transported by radiative flux up to the photosphere , where the optical depth @xmath53 . in our case @xmath54 or higher ,
therefore the trapping radius is very close to the photospheric radius and we will ignore in the following the radiative layer . the optical depth for electron scattering @xmath55 is computed with a thomson opacity @xmath56 , that we scale with the value appropriate for a flow composed solely of @xmath8-particles .
the electron density is , in fact , uncertain : it depends mainly on the initial composition of the wind at @xmath22 , which includes @xmath8-particles and free baryons . for a neutron - rich composition , @xmath57 .
the nucleosynthesis in the wind does not change the free electron density , since temperatures are high enough for the recombined helium to be fully ionized .
the trapping radius then reads : r_tr = 1.8 10 ^ 11 _
0.2 m_w_-3 t_w_-1 ^ 2/3 t_hr^-5/3 cm .
[ eq : rtr ] conservation of energy @xmath58 allows us to solve for the central temperature , @xmath59 and from eq .
[ eq : trt ] we can derive the temperature at the trapping radius t_tr(r_tr , t)1.5 10 ^ 5 ( ) t_hr^25/36 k. [ eq : txwind ] the emission from the trapping radius of the wind becomes harder with time while the luminosity , @xmath60 , decreases , l_tr(r_tr , t ) 2 10 ^ 40 ^2_0.3 ^ 2/3 _ 0.3 ^ -2/3 m_w_-3 ^ 1/3 t_w_-1 ^ 2/9 t_hr^-5/9 ergsec^-1 .
[ eq : lxwind ] when @xmath61 , t_x 14.9 ( ) ^3/5 t_w_-1 ^ 2/5 days , [ eq : tx ] the thermal emission has a temperature of t_x(,t_x ) 9 10 ^ 6 ( ) ^1/4 k , [ eq::tx ] and a luminosity l_x ( , t_x ) 7.5 10 ^ 38 _ 0.3 ^ 2 _ 0.2 ^ -1 ergsec^-1 .
[ eq : lx ] we note that in eq .
[ eq::tx ] and eq .
[ eq : lx ] the only dependences are on the initial radius , the terminal velocity and the opacity .
the dependence is particularly weak for @xmath62 because , when @xmath61 , the accretion rate is set only by the size of the launching region and by the opacity , ( see eq . [ eq : rtr ] ) .
therefore @xmath63 .
we stress the important role of the wind composition , equivalently of the electron fraction . for an extreme proton - to - neutron ratio of @xmath64
, @xmath65 and the x - ray emission ( @xmath66 k ) starts at @xmath67 with a luminosity @xmath68 erg / sec .
the emission from the wind will switch off when the whole energy supplied by the wind can be accreted ( @xmath69 ) .
this is a long time of the order of @xmath70 months .
however , when @xmath71 drops below @xmath72 , the x - ray emission is likely to be undetectable .
this happens for @xmath73 a few .
another scenario may be envisaged where a bound atmosphere forms around the central object .
this can happen , if the outflowing gas retains the same amount of energy per unit mass that the eccentric tail had .
this gas would still be bound to the central bh : it would start expanding from @xmath22 nearly isotropically until it reaches a radius @xmath74 , where its internal energy is @xmath75 half its potential energy .
after a few seconds , this inflated gas cloud has a nearly constant mass @xmath76 and radius @xmath74 , since most of the mass @xmath77 is injected around @xmath78 .
we can estimate the radius @xmath74 through @xmath79 , where @xmath80 is the semi - major axis of the particle orbits in the eccentric tail .
since the distribution of specific orbital energy @xmath81 with mass is constant @xmath82 , where @xmath83 is the extra energy gained by @xmath84 via the tidal torque , we can solve the integral and find @xmath85 .
we conclude that @xmath74 is of the same order as @xmath22 .
this reflects the fact that most of the mass is at small @xmath80 .
we choose to parametrize @xmath86 .
we can then calculate the cloud s mean properties .
its mean density is @xmath87 and the temperature can be derived equating its internal energy density ( in radiation and gas ) with @xmath88 .
radiation pressure is @xmath89 times higher than gas pressure and temperature decreases linearly with the cloud radius , @xmath90 ( while the pressure ratio remains constant ) .
the gas cloud is in hydrostatic equilibrium , since it changes its properties on a time scale @xmath91 , much longer than the dynamical timescale @xmath92 .
the rotational energy may be neglected , being a factor @xmath93 times smaller than the internal energy .
on timescales of interest , the cloud does not deflate via radiative losses , since the diffusion time is very long , @xmath94 .
the merged object is thus surrounded and obscured by a persistent source , emitting at the eddington limit l_ph = l_edd= = 6.2 10 ^ 38 _
0.2 ^ -1 erg sec^-1 , [ eq : ledd ] at a temperature @xmath95 .
after a few minutes , the photons escaping from the wind are in the ultraviolet band and after an hour or so in the extreme ultraviolet ( euv ) . this emission is strongly absorbed and unlikely to be observed . at later times
@xmath96 , however , the emission should peak in soft x - ray , with a luminosity @xmath97 ( eqs .
[ eq : lx ] and [ eq : ledd ] ) and a thermal spectrum with temperature @xmath98 ( eq . [ eq::tx ] ) . in the case that an opaque cloud surrounds the merged object , we have comparable luminosity , emitted at a temperature that depends on the extension or the atmosphere : for simplicity , we consider here the case in which the emission is at @xmath99 , corresponding to @xmath100 cm . the main difference with the wind case
is that this emission should be persistent . under favorable environmental conditions
, the emission may be observable .
if the merger occurs in a galactic halo or even in the intergalactic medium , absorption should be moderate .
moreover , those locations may not be polluted by contaminating soft x - ray sources .
finally , compact - object mergers should not be accompanied by a bright supernova explosion , eliminating another possible co - located x - ray source .
an eddington luminosity yields an unabsorbed flux at redshift @xmath101 of @xmath102 where we have approximated the luminosity distance at redshift @xmath101 as @xmath103 .
simulating the response of different current and future instruments , allows us to determine the expected count rate as a function of redshift . assuming @xmath104 ( galactic and intrinsic to the host galaxy ) and a black body spectrum , we get a count rate @xmath105 where @xmath106 for _ xmm _ , @xmath107 for _ chandra _ , @xmath108 for _ xeus _ and @xmath109 for _ con - x_. fig .
1 shows that x - ray detection is most likely to be feasible with the next generation of instruments . the proposed missions _
con - x _ and _ xeus _ will be able to collect @xmath110 cts or more in a @xmath111 sec exposure from mergers occurring as far as @xmath112 and @xmath113 respectively .
the local merger rate of ns - ns is estimated to be @xmath114 per milky way galaxy @xcite .
if half of those systems eventually merge in the halo , their rate is @xmath115 , where a number density of @xmath116 as been assumed @xcite . combining the ns - ns merger rate and the instrument observable volumes
, we predict that _ con - x _ could observe @xmath117 while the expected rate for @xmath118 is @xmath119 . in order to detect the x - ray emission ,
it is necessary to localize the merger . in the following , we discuss two possibilities .
coalescence of compact objects ( especially ns - ns ) are possible candidates as progenitors of sgrbs ( see * ? ? ?
* for a review ) .
therefore , in principle , a binary coalescence can be localized via a short burst , though there may be limitations .
first , the local observed rate is estimated to be @xmath120 times smaller than the local rate of compact object mergers .
the discrepancy is mostly credited to the geometrical beaming of the burst jet .
moreover , the _ observed _ redshifts range typically between @xmath121 .
however , the distances of these sources could only be measured for a handful of cases in the last few years . despite this
, it is reasonable to assume that short grbs should also explode closer to us , if they are indeed produced by ns - ns mergers , and that some selection effects are preventing us from measuring their redshifts .
when a short grbs with @xmath122 should be localized , the x - ray emission from the wind or the atmosphere could be brighter than the x - ray afterglow around two weeks later .
the gravitational wave signal is a more promising signpost for mergers .
this is because this signal should be associated with any coalescence of compact objects , unlike the beamed @xmath1-ray emission from sgrbs .
an instrument such as _ advanced ligo _ should be able to detect mergers of two neutron stars to a distance of @xmath123 : this implies a detection rate for _ xeus _ and _ con - x _ of @xmath124 . advanced _
ligo _ will be , in fact , more sensitive to ns - bh binaries , which should be visible up to @xmath125 .
since the galactic merger rate for these systems is @xmath126 per galaxy @xcite , _ xeus _ is expected to detect @xmath127 such mergers per year , while _ con - x _ less than @xmath128 per year .
the main limitation seems to be how accurately the merger position could be localized .
current estimates suggest that a network of non - collocated advanced interferometers such as advanced _ ligo _ , advanced _
virgo _ and _ lcgt _ , @xcite)will be able to detect inspiraling binaries at the redshifts of interest and localize them at a degree level @xcite .
this is enough for an optical but not for an x - ray follow - up .
however , * with a solid angle error * ten times smaller , we can identify a region of the sky with only one local galaxy with redshift @xmath129 ; for a galaxy at @xmath130 , the localization error should be , instead , a few hundreds times smaller .
the source distance maybe be obtained directly from the gravitational signal ( abbot at al .
this would greatly help the search for the x - ray fallback signal .
in this _ letter _ , we investigate the possible fate of fallback matter associated with mergers of compact objects , where a disc is formed by disruption of a ns .
matter flung to highly eccentric orbits will eventually come back to the disc at a super - eddington rate , converting its kinetic energy into heat via shocks , and will be unable to cool by neutrino emission .
contrary to previous claims , we think that this implies that fallback matter _ can not _
accrete all the way to the central object and be responsible for the late energy injections observed in grb afterglows .
rather , the fallback matter is likely to be blown off the disc plane , leading to the formation of a radiation - driven wind or a bound atmosphere . for the wind case
, we have analytically calculated the time evolution of the temperature and luminosity at the trapping radius : while the luminosity decreases ( eq . [ eq : lxwind ] ) , the wind photosphere becomes hotter ( eq . [ eq : txwind ] ) . at first
, the emission is in the euv band and absorption will likely prevent us from observing it .
after one or two weeks , the emission finally peaks in the soft x - ray band and the wind activity can be observed .
the bound cloud is radiation pressure - dominated and emits at the eddington limit in soft x - rays , if the atmosphere does not extend much further than @xmath131 cm .
we note that our estimates of luminosities are conservative : factors such as a smaller electron fraction in the ejected plasma and moderate geometrical beaming can substantially increase the expected luminosity .
we also discuss detection prospects for this delayed x - ray activity .
our inspection indicates that only in fortuitous circumstances could the x - ray emission be detected with current instruments , while the planned missions ( such as con - x and xeus ) have a better chance 5.1 .
then , the main limiting factors will not be the x - ray detector capability , but rather the tool for localizing the merger 5.2 .
on the one hand , short @xmath1-ray bursts can be easily detected and localized in the whole volume where instruments like _ con - x _ and _ xeus _ can observe the x - ray emission ; however , they are estimated to occur at a rate that is @xmath120 times smaller than the rate at which compact binaries merge . on the other hand , the planned advanced gravitational wave interferometers should be able to detect a signal from any such a merger but within cosmic distances smaller than the maximum distance that _ con - x _ and _ xeus _ can reach .
moreover , x - ray follow up would require better localization precision than currently estimated .
the net result is that between a few to a few tens of detections per year are expected by _
xeus _ with a follow - up of a short grb . assuming sufficiently good localization , re - pointing
after a gravitational signal detection can result in @xmath132 wind detection per year from ns - ns mergers , for both _ con - x _ and _ xeus_. furthermore for _ xeus _ , there is the exciting possibility to observe x - ray emission from bh - ns mergers : @xmath133 event per year .
the x - ray emission from these sources should also be brighter than from a ns - ns mergers , since the mass of the central bh could be much larger .
the above rates , however , should be taken as indicative of upper limits .
we have not taken into account selection effects such as background / foreground sources and the fact that not all bh - ns and ns - ns mergers seem to lead to an accreting system ( e.g. , * ? ? ?
* ; * ? ? ?
moreover , in some cases , the x - ray afterglow from the burst could outshine the wind emission .
nevertheless , the possibility to get information on mergers of compact objects from electromagnetic signals remains , and it could bring important understanding of the physics of these systems . finally , our findings have implications for interpreting late time activity observed in grb afterglows .
we consider unlikely that fallback matter can be held responsible , since most of the mass is blown away .
even if @xmath134 of this matter can accrete all the way to the hole , it is very unlikely that it could produce the observed flares , which have an energy ( @xmath135 ergs ) comparable to that of the prompt emission ( e.g. * ? ? ?
this would require that the eccentric tail is far more massive than the main disc ( contrary to what is observed in simulations ) or that the efficiency in converting accreted mass to energy is somehow strongly enhanced in the late fallback accretion .
these arguments also apply to the late accretion from the main disc , which is highly super - eddington @xcite .
we thus conclude that , in general , standard late time accretion is unlikely to account for the phenomena , like flares and plateaux , observed in grb afterglows .
the authors acknowledge useful discussions with k. belczynski , p. bender , p. armitage and e. quataert .
emr acknowledges support from nasa though chandra postdoctoral fellowship grant number pf5 - 60040 awarded by the chandra x - ray center , which is operated by the smithsonian astrophysical observatory for nasa under contract nasa8 - 03060 .
mcb acknowledges support through nasa astrophysics theory grant nng06gi06 g .
abbot , b. , and the ligo scientific collaboration , 2008 , phys . rev .
d 77 , 062002 belczynski k , taam r.e .
, kalogera v. , rasio f.a . , bulik , t. , 2007 , apj , 662 , 504 belczynski k , oshaughnessy r. , kalogera v. , rasio f.a . , taam r.e , bulik t. , 2008 , apj , 680 , 129 blinninkov s.i . ,
novikov i.d . ,
perevodchikova t.v . ,
polnarev a.g . , 1984 , soviet astron .
lett . , 10 , 177 campana s. et al . , 2006 , a&a , 454 , 113 chen w - x , beloborodov a. m. , 2007 , apj , 657:383 eichler d. , livio m. , piran t. , schramm , d. n. , 1989 , nat , 340 , 126 faulkner a. j. et al . , 2005 , apj , 618 , 119 faber j. a. , baumgarte t. w. , shapiro s. l. , taniguchi k. , rasio f. a. , 2006 , phys .
d , 73 , 4012 oshaughnessy r. , belczynski k. , kalogera v , 2008 , apj , 675 , 566 kazuaki k. et al , 2006 , class .
quantum grav . , 23 215 kim c. , kalogera v & lorimer d. r. , 2006 , proceedings of `` a life with stars '' , ed .
ralph wijers , amsterdam lee w. h. , ramirez - ruiz e. , 2006 , njph , 9 , 17 metzger b. d. , piro a. l. , quataert e. , 2008 , apj sub .
http://arxiv.org/abs/0805.4415 nakar e. , 2007 , phr , 442 , 166 paczyski b , 1991 , aca , 41 , 257 popham r. , woosley s. e. , fryer c. , 1999 , apj , 581,356 phinney e. s. , 1989 , iaus , 136 , 543 rosswog s. , speith r. , wynn g. a. , 2004 , mnras , 351 , 1121 rosswog s. , 2007 , mnras , 376 , 48 rosswog s. , 2005 , apj , 634 , 1202 ruffert m. , janka h - t , takahashi k. , schaefer g. , 1997 , a&a , 319 , 122 shibata m. , taniguchi k. , uryu , k. , 2003 , phys .
d , 68 , 4020 sylvestre , j. , 2003 , apj , 591 , 1152 | when double neutron star or neutron star - black hole binaries merge , the final remnant may comprise a central solar - mass black hole surrounded by a @xmath0 torus .
the subsequent evolution of this disc may be responsible for short @xmath1-ray bursts ( sgrbs ) .
a comparable amount of mass is ejected into eccentric orbits and will eventually fall back to the merger site after @xmath2 seconds . in this _ letter _
, we investigate analytically the fate of the fallback matter , which may provide a luminous signal long after the disc is exhausted .
we find that matter in the eccentric tail returns at a super - eddington rate and is eventually ( @xmath3 sec ) unable to cool via neutrino emission and accrete all the way to the black hole .
therefore , contrary to previous claims , our analysis suggests that fallback matter is _ not _ an efficient source of late time accretion power and is unlikely to cause the late flaring activity observed in sgrb afterglows .
the fallback matter rather forms a radiation - driven wind or a bound atmosphere . in both cases ,
the emitting plasma is very opaque and photons are released with a degraded energy in the x - ray band .
we therefore suggest that compact binary mergers could be followed by an x - ray renaissance " , as late as several days to weeks after the merger . this might be observed by the next generation of x - ray detectors
. black hole physics accretion , accretion discs |
the small magellanic cloud ( smc ) is host to a large population of @xmath350 high - mass x - ray binaries ( hmxbs ) , comparable to the number known in the galaxy ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. however , unlike the galactic population , all of these systems but one are be / x - ray binary ( be - xrb ) systems , with the exception ( smc - x1 ) containing a supergiant companion instead @xcite .
the wing of the smc is the region linking the eastern side of the smc to the bridge . despite its lower content of gas , dust , and stars
, it had a major star forming event @xmath311 myr ago and appears , as expected , deficient in be - xrbs @xcite .
note however that the coverage of the smc wing by x - ray observations has been sparser than that of the bar .
each new discovery of a be - xrb in the smc wing is therefore particularly noteworthy ( e.g. * ? ? ?
we have conducted _ chandra _ and _ xmm - newton _ observations of the far eastern region of the wing , concentrating on the star forming region ngc602 .
the full description of these observations will be presented in a forthcoming paper ( oskinova et al .
, in preparation ) . in this letter , we concentrate on one bright x - ray source located near ngc602 in projection and coinciding with the emission - line star 2dfs3831 ( r.a .
= @xmath4 , dec .
@xmath5 , j2000.0 ) , which is revealed as a be - xrb . based on the results of section [ timing ] , throughout this paper
we will refer to the x - ray source as sxp1062 following the nomenclature of @xcite for smc x - ray pulsars .
the stellar population of ngc602 ( which comprises three star clusters : ngc602a , b , and c ) has been studied in detail .
in addition to very old stars of 6 - 8 gyr , likely the smc field population , the ngc602 clusters contains young stars of @xmath34 - 5myr @xcite .
the significant number of young stellar objects detected in the ngc602a cluster @xcite suggests that star formation is still ongoing . as for what initially triggered the formation of these clusters , the low density environment in the wing of the smc hints at an additional mechanism acting along with gravitational collapse .
ngc602 is located at the intersection of three h@xmath1i shells , and @xcite suggested that its formation is the result of the interaction of two expanding shells @xmath37 myr ago .
@xcite proposed instead that star formation was possibly induced by encounters with the large magellanic cloud or the milky way .
ngc602 is associated with sgs - smc1 , the only h@xmath0 supergiant shell known in the smc .
the discovery of a be - xrb in this area ( only @xmath37@xmath6 to the west of ngc602a ) provides important clues to the star formation history in the wing and the evolution of large - scale structures in the interstellar medium .
this letter is organized as follows .
we describe the x - ray observations and analysis in section [ x - ray ] . in section [ optical ] ,
we present the spectroscopy of the optical counterpart .
we discuss the properties and environment of sxp1062 in section [ discussion ] and present our conclusions in section [ conc ] .
the x - ray data were obtained with the acis - i camera on the _ chandra _ x - ray observatory and with the epic cameras onboard _ xmm - newton_. the _ chandra _ observations consisted of 11 separate exposures acquired between 2010 - 03 - 31 and 2010 - 04 - 29 ( effective exposure time of 290.7ks ) , while 4 separate exposures were obtained with _ xmm - newton _ between 2010 - 03 - 25 and 2010 - 04 - 12 ( epic - pn effective exposure time of 162.5ks ) . the data were reduced using the most up - to - date versions of the corresponding data reduction software .
the x - ray source cxoj012745.97 - 733256.5 ( = sxp1062 ) coinciding with the optical emission - line star 2dfs3831 is detected in each of these 15 exposures . in the _ chandra _ observations ,
the positional uncertainty ( 1@xmath7 ) of the x - ray source is @xmath8 .
the source was also detected during an _ xmm - newton _
slew on 2009 - 11 - 16 .
interestingly , the source was not detected in the _
rosat all sky survey _ , although at its present luminosity it would have been bright enough at an expected _ rosat _ pspc count rate of 0.01 counts s@xmath9 .
figure[bestfit_spec ] shows the combined background - subtracted _ chandra _ and _ xmm - newton _ spectra of sxp1062 .
spectra were extracted from a 128 circular region and the background was extracted from a concentric annular region of radii 128 and 192 which was free of sources .
likewise , the _ xmm - newton _ spectra were extracted from a 32 circular region and the background from 4 adjacent circular regions free of sources with radius 45 . [ cols="<,^,^,^,^,^,^,^,^ " , ] as a first step the background - subtracted _ chandra _ and _ xmm - newton _ spectra of sxp1062 were simultaneously fitted using a simple absorbed power - law model .
the best - fit parameters , listed in table[tab : fit ] , imply an absorbed flux in the energy range 0.2 - 12.0 kev of @xmath10=1.8@xmath1110@xmath12 erg @xmath13 s@xmath9 .
this corresponds to an intrinsic x - ray luminosity in this same energy range of @xmath14=6.9@xmath1110@xmath15 erg s@xmath9 assuming a distance modulus of 18.7 appropriate for the wing ( e.g. * ? ? ?
* ) , or @xmath16=8.2@xmath1110@xmath15 erg s@xmath9 for the ` standard ' smc distance modulus of 18.9 ( e.g. * ? ? ?
the photon index @xmath17 of @xmath30.75 is typical of x - ray pulsar binaries ( e.g. * ? ? ?
* ) and a signature of accretion onto a strongly magnetized neutron star .
the simple power - law model suggests some emission excess below 1 kev and at high energies . as a next step
, we added a blackbody component , which statistically improved the fit ( see table[tab : fit ] ) .
the observed x - ray flux is similar to that of the previous model , but the x - ray luminosity decreases by @xmath18 .
we also tried a model for which a thermal plasma component with smc abundances is included instead of a blackbody . similarly to the blackbody , the addition of a thermal component to the power - law model improves the fit . in this case , the intrinsic x - ray luminosity is only marginally greater by @xmath19 compared to the simple power - law model .
finally , the best fit is achieved by adding simultaneously a thermal and a blackbody component to the power - law ( see table[tab : fit ] and figure[bestfit_spec ] ) . the flux for this model ( pbt in table [ tab : fit ] )
is @xmath20=1.66@xmath2110@xmath12 erg @xmath13 s@xmath9 , corresponding to an intrinsic luminosity @xmath226.3@xmath2310@xmath15 erg s@xmath9 for a distance modulus of 18.7 .
the column densities are well constrained by spectral fitting , albeit the values slightly differ depending on the model . using the best - fit values , we note that the h column densities in table[tab : fit ] imply @xmath24 mag for the @xmath25/@xmath26 ratio of @xmath27@xmath13 , i.e. @xmath28 assuming @xmath29 .
this range is in good agreement with the @xmath30 value computed by comparing the value @xmath31 of 2dfs3831 @xcite with the intrinsic value @xmath32 of a b0 - 0.5iii star @xcite given the uncertainties on the photometry and spectral type ( see section [ optical ] for the optical spectroscopy of 2dfs3831 ) .
the photon arrival times were corrected for the solar system barycenter .
we searched for pulsations in the x - ray light curves in the soft and hard energy bands ( 0.4kev1.5kev , 2.0kev7.2kev ) and a total energy band ( 0.4kev7.2kev ) , using fast fourier transform and light - curve folding techniques as implemented in the timing analysis software _
xronos_. figure[power ] shows the inferred _ chandra _ and _ xmm - newton _ power density spectra in the soft and hard energy bands with a clear peak at a frequency of @xmath33hz ( p=1062s ) . this coherent x - ray pulse period establishes the source as a binary x - ray pulsar .
the pulse profiles folded with this period are shown in figure [ pulse ] .
the light - curves used to compute the _ xmm - newton _ power spectrum in the hard energy band , the _ xmm - newton _ power spectrum in the soft energy band , and the _ chandra _ power spectrum in the hard energy band were binned by respectively 5s , 200s , and 200s .
the 5s bin time for one of these was chosen to make sure that a shorter pulse was not missed .
no pulse was found in the soft band _
observations , possibly because the count rate is too low . because of the softer response of the epic - pn camera , the pulse in the soft band is the most obvious in the light - curves obtained with this instrument .
note that in addition to the pulse period , x - ray variability was also detected at a level of @xmath34 peak - to - peak on a timescale of several days in both _
xmm - newton _ and _ chandra _ light - curves .
given that this variability does not appear regular and that the time sampling is very sparse , we do not analyse these variations further .
from its x - ray position with subarcsecond precision from _ chandra _ images , the optical counterpart to sxp1062 is identified as 2dfs3831 @xcite .
we observed this star with the vlt flames instrument @xcite on 2010 october 25 as part of a spectroscopic survey of massive stars in ngc602 complementing our x - ray observations .
spectra were obtained in the medusa - fibre mode of flames using the lr02 setting of the giraffe spectrograph ( 39604564 , r=7000 , e.g. * ? ? ? * ) .
five pairs of 1800s exposures were obtained .
the standard data processing ( bias subtraction , fibre location , summed extractions , division by flat - field , wavelength calibration ) was done using the eso common pipeline library flames reduction routines ( v.2.8.7 ) .
additionally , heliocentric correction and subtraction of a median sky spectrum were performed ( for details see e.g. * ? ? ?
the spectra from all the exposures were then normalized and merged .
the 2df spectra of this star were also retrieved .
these two spectra cover the regions from @xmath33800 to 4900 and @xmath36000 to 7000 , with resolving powers of 1500 and 2500 respectively @xcite .
figure [ opt ] shows the vlt flames and 2df spectra of 2dfs3831 .
the vlt flames spectrum has been smoothed and rebinned to an equivalent resolving power of r=4000 .
a weak he@xmath1ii @xmath354542 absorption line is visible in the vlt flames spectrum and there is a hint of a weak he@xmath1ii @xmath354686 line in the 2df spectrum , but he@xmath1ii @xmath354200 is absent , suggesting a spectral type around b0 - 0.5 following the classification adopted by @xcite .
several characteristics attributable to a circumstellar disc indicate that 2dfs3831 is a classical be star .
fe@xmath1ii @xmath354179 and fe@xmath1ii @xmath354233 emission lines typical of some early be stars ( e.g. * ? ? ?
* ) are visible in the vlt flames spectrum .
the h@xmath0 emission seen in the 2df spectrum ( see the right panel in fig.[opt ] ) is relatively strong with an equivalent width ( ew ) of -23 .
there is significant h@xmath36 emission , plus emission in the core of the other balmer lines ( double - peaked in the vlt flames spectrum ) and apparent infilling of the he@xmath1i absorption lines .
two micron all sky survey ( 2mass ; * ? ? ?
* ) jhk@xmath37 photometry also indicates a clear infrared excess when compared with expected colors for a b0 - 0.5 spectral type @xcite .
the double - peaked emission in the balmer lines is not due to oversubtraction of nebular features present in the median sky spectrum .
also , given the relative weakness of the nebular emission lines in the mean sky spectrum , the emission in the core of the balmer lines is most likely dominated by circumstellar material and not by nebular contamination .
the observed magnitude ( v=14.36 , b=14.32 ; * ? ? ? * ) and a range of reasonable extinction estimates ( including both the interstellar and circumstellar components ) lead to an absolute magnitude consistent with a b0 - 0.5 giant @xcite .
we thus determine the spectral type of 2dfs3831 as b0 - 0.5(iii)e+ , where ` + ' signifies the presence of fe@xmath1ii .
this corresponds to a typical effective temperature @xmath38k and an evolutionary mass @xmath39 at smc metallicity ( cf .
we estimated the radial velocity of 2dfs3831 from the vlt flames spectrum by fitting gaussian profiles to the wings of h@xmath40 , h@xmath41 , h@xmath42 , he@xmath1i @xmath354143 , and he@xmath1i @xmath354388 .
the average of all measurements is 167 with a standard deviation of 13 .
this is consistent with the mean velocity of massive stars in the smc @xcite , and also with the gas velocities measured by @xcite across the n90 h@xmath1ii region ionized by ngc602a .
thus , we do not see any evidence that the hmxb acquired a high space velocity following the supernova explosion .
sxp1062 is only the third smc x - ray pulsar with a spin period larger than 1000s @xcite .
these slowly rotating pulsars are particularly interesting because they represent a challenge for the theory of spin evolution of a neutron star in a close binary system ( e.g. * ? ? ?
according to the corbet diagram for smc be - xrbs relating spin and orbital period @xcite , we expect sxp1062 to have a binary period of @xmath3300 days .
such a long orbital period is also expected from the @xmath43-ew(h@xmath0 ) diagram @xcite , from which our measured ew(h@xmath0 ) suggests an orbital period of @xmath3100 days .
note however that the maximum ew(h@xmath0 ) of the system , which probes the maximum size of the disc ( and indirectly the orbital period ) , could be higher than our instantaneous measurement , so this period estimate should be taken as a lower limit .
sxp1062 shares many characteristics with the class of persistent be - xrbs ( e.g. * ? ? ?
* ) : a relatively low x - ray luminosity @xmath44ergs@xmath9 , a slowly rotating pulsar with @xmath45s , a relatively flat light curve with sporadic and unpredicted increases in intensity by less than an order of magnitude , and a lack of iron lines at @xmath36.4kev indicative of small amounts of material in the vicinity of the neutron star .
a thermal excess of blackbody type , with a high temperature ( @xmath46kev ) and a small emission area ( @xmath47 km , consistent with a hot spot at the polar cap of the neutron star ) has recently been suggested as a common feature of persistent be - xrbs @xcite .
such a component is also identified in the x - ray spectrum of sxp1062 , for which we infer a size of @xmath48 km for the blackbody source from our best - fit models ( section[xspec ] ) .
a shell nebula is detected around sxp1062 in the magellanic cloud emission - line survey ( mcels ; * ? ? ?
* ) h@xmath0 image and in the higher resolution ctio 4 m mosaic @xcite h@xmath0 image ( see figure [ mcels ] ) .
this shell is also detected in the mcels [ o@xmath1iii ] image , but hardly in [ s@xmath1ii ] .
we adopt below a distance modulus of 18.7 .
the shell is not uniform , with radii varying from 75@xmath49 at the sharp rim on the northeast to 94@xmath49 at the diffuse edge on the southwest .
the peak surface brightness of the northeast rim is @xmath50 ergs @xmath13 s@xmath9 arcsec@xmath51 , corresponding to an emission measure of 30 @xmath52 pc .
the width of the rim suggests that the shell thickness ( @xmath53 ) is 510% of the shell radius ( @xmath54 ) .
the longest emitting length at the shell rim is @xmath55^{1/2}$ ] ; thus , the rms density in the shell is [email protected] h - atom @xmath57 .
the gas mass in the shell is 250@xmath56100 @xmath58 .
this large mass indicates that the shell gas must be dominated by interstellar material .
the shell morphology resembles supernova remnants ( snrs ) in the magellanic clouds .
the x - ray images also suggest that diffuse x - ray emission possibly associated with a snr may be present in the vicinity of sxp1062 ( oskinova et al .
, in preparation ) .
as the shell mass is much larger than the typical supernova ejecta mass , we assume that the shell is a snr in the sedov phase .
the kinetic energy in the shell would be @xmath330% of the supernova explosion energy . adopting a canonical explosion energy of 10@xmath59 ergs ,
the current shell kinetic energy is @xmath60 ergs , and implies a shell expansion velocity of 350@xmath56100 km s@xmath9 and an age of 0.4(@xmath61)=@xmath2 yr .
this age is much larger than the cooling timescale of this low density gas , justifying an adiabatic shock for the sedov phase . the pre - shock interstellar gas density , 1/4 the shell density , is @xmath62 h - atom @xmath57 , consistent with the low density expected in the smc wing .
the bright [ o@xmath1iii ] emission can be easily produced by a 350 km s@xmath9 shock ( cf .
the ionizing flux of the b00.5iii star can easily photoionize the shell gas and its surrounding pre - shock medium ; furthermore , the diffuse [ o@xmath1iii ] emission to the northeast exterior of the shell indicates the existence of a harsh radiation field ; therefore , the ionization stage of the pre - shock medium may be too high to produce strong post - shock [ s@xmath1ii ] emission . we thus conclude that the shell nebula detected around sxp1062 is most likely a snr .
the extent of the shell , which has not reached the ngc602 region , and its young kinematic age suggest that this supernova event did not trigger the formation of ngc602 .
image and mcels h@xmath0 , [ s@xmath1ii ] , and [ o@xmath1iii ] images of a region centered on the position of 2dfs3831=sxp1062 and showing the shell nebula detected around the target.,width=302 ]
we reported the discovery of a be - xrb , sxp1062 , containing one of the slowest rotating x - ray pulsars in the smc .
we suggest that the shell nebula detected around this object is a snr .
its estimated kinematic age of @xmath2 yr is probably the age of the pulsar .
this is , to our knowledge , the first discovery of a pulsar associated with a supernova remnant in the smc .
this study is based on observations obtained with _ xmm - newton _ , an esa science mission with instruments and contributions directly funded by esa member states and nasa , and the _ chandra _ science mission .
the software provided by the _
chandra _ x - ray center ( cxc ) in the application package ciao and by the _ xmm - newton _ in the package sas was used .
this research has made use of nasa s astrophysics data system service and the simbad database , operated at cds , strasbourg , france .
we thank frank winkler for providing the smc mcels images .
vhb acknowledges support from supa and nserc .
lmo . acknowledges support from the dlr grant 50or0804 .
mag is supported by the micinn grant aya2008 - 01934 that includes feder funds .
w.s . acknowledges support from the daad grant a/10/95420 .
the dfg grant os292/3 - 1 supported the project workshop .
yhc , jsg , and rag acknowledge the support of nasa grant sao go0 - 11025x . | we report on a new be / x - ray pulsar binary located in the wing of the small magellanic cloud ( smc ) .
the strong pulsed x - ray source was discovered with the _ chandra _ and _ xmm - newton _ x - ray observatories .
the x - ray pulse period of 1062s is consistently determined from both _
chandra _ and _ xmm - newton _ observations , revealing one of the slowest rotating x - ray pulsars known in the smc .
the optical counterpart of the x - ray source is the emission - line star 2dfs3831 .
its b0 - 0.5(iii)e+ spectral type is determined from vlt - flames and 2df optical spectroscopy , establishing the system as a be / x - ray binary ( be - xrb ) .
the hard x - ray spectrum is well fitted by a power - law with additional thermal and blackbody components , the latter reminiscent of persistent be - xrbs .
this system is the first evidence of a recent supernova in the low density surroundings of ngc602 .
we detect a shell nebula around 2dfs3831 in h@xmath0 and [ o@xmath1iii ] images and conclude that it is most likely a supernova remnant . if it is linked to the supernova explosion that created this new x - ray pulsar , its kinematic age of @xmath2 yr provides a constraint on the age of the pulsar .
[ firstpage ] x - rays : binaries stars : emission - line , be magellanic clouds . |
pulsars are widely believed to be fast rotating neutron stars , which are compact objects with typical radius @xmath4 cm and mass @xmath5 .
pulsars usually have relatively stable pulse profiles and rotational periods , due to which they can even be used as unique high - precision clocks in experimental astrophysics . however , sometimes there are some timing variabilities observed in pulsars , such as glitches ( wang et al .
2000 ; yuan et al . 2010 ; espinoza et al .
2011 ; manchester & hobbs 2011 ; yu et al . 2013 ) , micro - glitches ( cognard & backer 2004 ; mandal et al .
2009 ) , pulse profile changing ( rankin 1986 ; burgay et al . 2005 ; poutanen et al . 2009 ; karastergiou et al .
2011 ; bilous et al .
2014 ) , pulse nulling ( deich et al . 1986 ; rankin & wright 2008 ; jones 2011 ; li et al .
2012 ) , and pulse drifting ( backer 1973 ; page 1973 ; ruderman 1976 ; esamdin et al . 2005 ; jones 2014 ) .
especially , neutron star glitches , characterized by a sudden increase and gradual relaxation of rotation frequency , are very interesting astrophysical phenomena and have been observed from many normal pulsars and magnetars ( kaspi et al .
2000 ; dib et al .
2009 ; livingstone et al .
2010 ; gavriil et al .
2011 ) , where magnetars are a type of pulsars with dipole magnetic fields significantly stronger than @xmath6 g ( usov 1992 ; ducan & thompson 1992 ; olausen & kaspi 2014 ) .
usually , the glitch is explained to be associated with sudden decoupling of the pinned vortex lines in the crustal neutron superfluid region ( anderson & itoh 1975 ; pines & alpar 1985 ; alpar & baykal 2006 ; warszawski & melatos 2011 ; haskell et al .
2012 ; warszawski et al . 2012
; chamel 2013 ) . in the normal steady state ,
the pinned superfluid is coupled to the rest of the neutron star .
but when the pinned superfluid is decoupled due to continuously increasing rotation lag , the angular momentum of the fast - rotating interior superfluid component will be transferred to the outer solid crust , leading to an observed glitch .
there are also some other models to explain the observed glitches , such as the platelet collapse model ( morley & schmidt 1996 ) , the superfluid r - mode instability mechanism ( glampedakis & andersson 2009 ) , the snowplow model ( seveso et al . 2012 ) , and the starquake model ( zhou et al . 2014 ) .
anti - glitches were also observed in pulsars , such as from 1e 2259 + 586 ( archibald et al .
2013 ) and 1e 1841 - 045 ( sasmaz mus et al .
anti - glitches could be generated by either an internal mechanism such as an impulsive angular momentum transfer between the superfluid region and the crust ( thompson et al .
2000 ) , or an external mechanism such as a sudden twisting of the magnetic field lines ( lyutikov 2013 ) or accretion of retrograde matter ( katz 2014 ; ouyed et al . 2014 ) . however , the anti - glitch of the magnetar 1e 2259 + 586 was very special , associated with a hard x - ray burst , which strongly challenges traditional glitch theories .
huang & geng ( 2014 ) proposed a completely different model to interpret this strange behavior . in their model , the sudden spin - down
is explained to be induced by collision of a small solid body with the central magnetar .
the associated hard x - ray burst and the decaying softer x - ray emission can be explained well .
recently , a sudden change in the spin - down rate of psr j0738 - 4042 was reported by brook et al .
the value of the spin - down rate was originally quite stable .
but after september 2005 , a significant reduction in the spin - down rate was observed .
this abrupt change was coincided with the appearance of a new component in the average pulse profile ( karastergiou et al .
as these timing and emission properties can not be explained by normal intrinsic pulsar processes , brook et al .
( 2014 ) argued that they were generated by an external mechanism and invoke an asteroid disrupted by the central neutron star .
however , we noticed that their calculations are not self - consistent .
here we show that accretion from tidally disrupted material is truly a possible explanation for the special behaviors observed in the spin - down rate and present a self - consistent calculation .
the structure of our paper is organized as follows .
we summarize the observational facts of psr j0738 - 4042 in section 2 .
the asteroid disruption model , including the tidal disruption and accretion processes , are described in section 3 .
we summarize our results in the final section with a brief discussion .
psr j0738 - 4042 is a radio - emitting neutron star with a rotational period of 0.375 s and spin - down evolution of @xmath7 @xmath8 ( brook et al .
2014 ) , where @xmath9 is the time derivative of the rotational frequency @xmath10 , which is gradually decreasing due to magnetic dipole radiation . as the target of a long - term monitoring campaign ,
psr j0738 - 4042 has been observed by researchers at the hartebeesthoek radio astronomy observatory in south africa from september 1988 .
psr j0738 - 4042 has also been observed by the parkes radio telescope in australia as part of the fermi timing programme since 2007 ( weltevrede et al .
psr j0738 - 4042 shares similar rotational properties with other isolated , middle - aged radio pulsars .
the pulse profile and timing property of psr j0738 - 4042 are originally stable , with @xmath7 @xmath8 ( see fig .
1 ) . in fig .
1 , since a smaller @xmath9 value means the pulsar is spinning - down relatively quickly , we call this initial stage as the fast spinning - down phase . however , from september 2005 , a dramatic change in the spin - down rate occurred ( brook et al . 2014 , also see fig . 1 for a schematic illustration ) .
this change was accompanied by a new radio component that drifted on the leading edge of the pulse profile , which was first noted by karastergiou et al .
the value of @xmath9 gradually rose up to a peak value of @xmath1 about 450 days later since september 2005 .
the amplitude of this change in the spin - down rate is about @xmath11 .
note that there are also some small variabilities during the rising phase . as marked in fig . 1
, we call this stage as the first pulse . after reaching the peak value ,
@xmath9 began to drop down slightly . in fig .
1 , this stage is marked by
the drop " . after dropping down to a value of @xmath12 , another significant change in the spin - down rate began .
the value of @xmath9 rose back to @xmath13 in about 300 days .
after some small turbulence , the spin - down rate finally arrived at another relatively stable phase , with a new value of @xmath14 ( brook et al .
we call this stage the slowly spinning - down phase , since a higher @xmath9 value indicates that the pulsar is spinning - down relatively slowly . as the change in the spin - down rate of psr j0738 - 4042 is accompanied by an emergent radio component that drifts with respect to the rest of the pulse profile , normal intrinsic pulsar processes can not explain these radio emission and timing features .
brook et al ( 2014 ) suggested that they are witnessing an encounter of the pulsar with an asteroid .
due to orbital perturbations and collisions , asteroids may migrate inwards and interact with the magnetosphere after being disrupted by the central pulsar , affecting the pulse profile and the rotational stability . according to brook et al .
( 2014 ) , the change in the spin - down rate was caused by mass supplied to the pulsar .
however , when calculating the total accreted mass , they firstly related the reduction in the outflowing charge density to the change in the spin - down rate
. then they multiplied the reduced charge density with the speed of light , the polar cap area , and the duration of the new spin - down state to get a mass of about @xmath15 g. they pointed out that this value is the mass of the asteroid encountering the central pulsar .
but in fact , the mass estimated by brook et al .
( 2014 ) is the reduction mass in the total outflowing plasma above the polar caps .
it can not be regarded as the mass of the accreted material , which itself is a kind of inflow .
thus it is obvious that their interpretation and calculations are not self - consistent . in this paper , to explain the behavior of the spin - down rate of psr j0738 - 4042 , we reconsider the tidal disruption and accretion processes in detail and give a self - consistent modeling .
interestingly , shannon et al ( 2013 ) showed that the timing variabilities of psr b1937 + 21 is consistent with the existence of an asteroid belt .
actually , the planetary and disk systems have been confirmed in some neutron stars , such as psr b1257 + 12 ( wolszczan 1994 ) , psr b1620 - 26 ( thorsett et al . 1999 ) and the magnetar 4u 0142 + 61 ( wang et al .
the mechanism of collision between small bodies and neutron stars has also been widely used to interpret transient x/@xmath16-ray events ( van buren 1981 ; livio & taam 1987 ; zhang et al .
2000 ; campana et al .
2011 ) . to explain the pronounced change in the spin - down rate of psr j0738 - 4042
, we invoke an close encounter between an asteroid and the central pulsar .
since the spin - down rate reduced significantly ( the @xmath9 value increased , see fig .
1 ) after september 2005 , the orbital angular momentum of the asteroid is assumed to be parallel to that of psr j0738 - 4042 .
firstly , the asteroid will be disrupted at the tidal disruption radius .
after the disruption , some fraction of the disrupted material will be ejected at high speed , while the rest is bound to the central pulsar ( rees 1988 ) .
when the bound material moves around the central pulsar , the gaseous debris will be partially accreted on to the surface of the central pulsar , changing the emission and pulse profile signatures .
initially , the bound orbit is highly eccentric .
as the material tends to stay in an orbit with the lowest energy for a given angular momentum , i.e. a circular orbit , the disrupted and bound debris will experience a process called the orbital circularization .
after the orbital circularization , a disk would be formed . from then on , the mass accretion rate tends to be somewhat constant before the exhaustion of the material in the transient disk .
we show that the evolution behavior of the spin - down rate of psr j0738 - 4042 can be explained by the accretion processes at different stages . due to strong gravitational force
, the asteroid that comes too close to the pulsar will be broken up when coming to the tidal disruption radius given by ( hills 1975 ) @xmath17 where @xmath18 is the pulsar mass and @xmath19 is the characteristic central density of the asteroid , which will be taken as @xmath20 for a typical homogeneous iron - nickel body ( colgate & petschek 1981 ) for simplicity in our calculations .
thus the tidal disruption radius can be calculated to be @xmath21 typically .
after the disruption , part of the disrupted material will be bound to the central pulsar . in the first flyby
, a small portion of the bound material will be accreted towards the central pulsar . during the accretion process
, the infalling material will exert its own force to generate the ram pressure .
because of the existence of the dipole magnetic field , there will also be a magnetic pressure at any given radius .
equating the two pressures , we can get the @xmath22 radius , which is also known as the magnetospheric radius , given by @xmath23 where g is the gravitational constant and @xmath24 is the mass accretion rate .
@xmath25 is the magnetic dipole moment of the pulsar and @xmath26 is the surface magnetic field , which can be estimated as @xmath27 g for psr j0738 - 4042 , with @xmath10 and @xmath9 being 2.667 @xmath28 and @xmath29 respectively ( see the fast spinning - down phase in fig .
1 ) . in our calculations ,
the disrupted material is assumed to be accreted and adhered to the central pulsar from the magnetospheric radius .
conservation of angular momentum then leads to @xmath30 where @xmath31 is the moment of inertia of the central pulsar , which is taken as a typical value of @xmath32 ( pizzochero 2011 ; hooker et al . 2013 ) , and @xmath33 is the mass of the accreted material .
@xmath34 is the velocity of the asteroid at the radius @xmath35 , assumed to be @xmath36 . through some simple derivation
we can further simplify equation ( 3 ) to @xmath37 to estimate the characteristic mass accretion rate , we firstly assume it to be constant .
@xmath38 is taken as @xmath11 , which is determined from observational data ( i.e. , the difference of @xmath9 in the fast spinning - down phase and in the first pulse phase , see fig .
1 ) . for typical parameters of @xmath39 and @xmath40 cm ,
the average mass accretion rate can be calculated as @xmath41 . in this case , the @xmath22 radius can be estimated as @xmath42 .
we argue that the first pulse phase beginning from september 2005 in fig .
1 is due to the initial accretion of the gaseous debris in the first flyby after the disruption .
based on observations , there is a drop in the @xmath43 diagram after the first pulse , as shown in fig .
1 . this is because , after passing the pericenter , the majority of the disrupted material will fly away from the central pulsar along the highly elliptical orbit .
it will result in a decrease of the mass accretion rate , leading to a drop of @xmath9 . in our calculations
, we assume the duration of the first pulse , which is about 700 days , as the initial orbital period ( @xmath44 ) of the bound debris .
according to the formula @xmath45 , the semimajor axis @xmath46 of the initial elliptical orbit can be calculated as @xmath47 cm , which is larger than the tidal disruption radius and the @xmath22 radius , indicating that our calculations are self - consistent . assuming that the initial orbital period is also roughly the time for the bound debris to return to the pericenter , we further have ( ulmer 1999 ; lu et al .
2006 ) @xmath48 where @xmath49 is the radius of the asteroid and @xmath50 is the pericenter distance . from this equation
, the pericenter distance can be estimated as @xmath51 cm .
when the disrupted material moves around the central pulsar in elliptical orbit with high ellipticity , there will be dissipative processes , e.g. collisions , shocks , viscous dissipation , etc ( shakura & sunyaev 1973 ) .
these processes will convert some of the energy of the ordered bulk orbital motion into internal energy , part of which will be radiated and therefore lost from the gas . thus the gas has to sink deeper into the potential of the central pulsar , orbiting it more closely .
this in turn requires the gas to lose angular momentum .
so most of the disrupted material will spiral towards the central pulsar through a series of elliptical orbits with continuously decreasing ellipticities .
the above process is called the orbital circularization . during the circularization
, some gas will move inwards due to the alpha viscosity process . in this study , as the corotation radius @xmath52 is smaller than the @xmath22 radius ( @xmath53 ) , the gaseous debris will firstly accumulate at the @xmath22 radius of the pulsar .
when the pressure exerted by the accreted material exceeds the magnetic pressure from the central pulsar , the gas will stream on to the poles of the pulsar along the magnetic field lines , changing the spin and emission properties of the central psr j0738 - 4042 .
after the circularization , an accretion disk would be formed around psr j0738 - 4042 . the observed relatively stable spin - down rate at the last stage in fig . 1 ( i.e. , the slowly spinning - down phase ) can be explained by the constant accretion from the disk . comparing @xmath9 in the fast spinning - down phase with that in the slowly spinning - down phase
, we again get the difference as @xmath54 .
this again gives the accretion rate as @xmath41 . during the accretion timescale , i.e. from september 2005 to right now , the total accreted mass should be @xmath55 g , which is consistent with the mass range of asteroids around neutron stars . in this study , we regard @xmath56 as the lower limit of the mass of the asteroid . the total time for the bound debris to complete the circularization process around the central compact object is ( ulmer 1999 ; lu et al . 2006 )
@xmath57 where @xmath58 is the number of orbits necessary for the circularization , usually ranging between 2 and 10 . for @xmath59 days
, we get @xmath60 days .
it is interesting to note that in the @xmath43 diagram ( fig .
1 ) , beginning from september 2005 , a total period of about 1800 days can be isolated , during which the curve shows noticeable variabilities before it finally enters the stable slowly spinning - down phase after about 2011 .
this time span is roughly consistent with the @xmath61 value that we derived , indicating that the circularization process was in progress in this period .
in this study , by invoking an asteroid disrupted and accreted by psr j0738 - 4042 , we show that the pronounced change of the spin - down rate can be reasonably explained . especially , the sudden initial reduction ( i.e. the first pulse in fig .
1 ) of the spin - down rate comes from the initial accretion when the asteroid flies by the central pulsar for the first time , while the subsequent drop ( the drop in fig .
1 ) is due to the flying away of the bound gaseous debris . the relatively stable spin - down rate at the last stage ( the slowly spinning - down phase in fig .
1 ) can be explained to be resulted from the constant accretion from the transient accretion disk formed after the orbital circularization . as suggested by brook et al .
( 2014 ) , we also expect the spin - down rate to return to its previous value when the material from the disrupted asteroid is exhausted . for a typical asteroid with a characteristic mass of @xmath62 ( with radius @xmath63 ) ,
the disrupted material will be entirely swallowed by psr j0738 - 4042 in a timescale of about @xmath64 yr .
however , notice that the mass range of the asteroid can extend from about @xmath65 to @xmath66 , so the exact accretion time is quite uncertain .
in the future , when the spin - down rate of psr j0738 - 4042 returns to its previous value , we can use the accretion timescale to constrain the mass of the asteroid .
huang & geng ( 2014 ) proposed an external mechanism to explain the unprecedented anti - glitch observed in the magnetar 1e 2259 + 586 . in their model
, the impact parameter is very small so that the solid body will collide with the neutron star before coming to the periastron .
as their collision process is very fast and violent , a glitch and an associated hard x - ray burst will be produced .
while in our case , the encounter of the small body with the neutron star does not lead to a direct collision .
it is a much gentler process , also greatly prolonged .
the change in the spin - down rate of psr j0738 - 4042 comes from the tidal disruption and accretion of the asteroid , which is much longer and smoother compared with that happens in a direct collision .
therefore mainly the spin - down rate of psr j0738 - 4042 is affected and there is no obvious glitch induced . in our calculations , the pericenter distance of the captured asteroid
is estimated as @xmath51 cm , which is smaller than the tidal disruption radius ( @xmath67 cm ) . in this case ,
named as an ultraclose passage by rees ( 1988 ) , the asteroid will be completely disrupted . as discussed by carter & luminet ( 1982 ) ,
such an asteroid is not only elongated along the orbital direction in the process , but also will undergo compression to form a short - lived pancake aligned in the orbital plane .
they defined a factor @xmath68 ( @xmath69 ) to derive some key parameters in the ultraclose passage case .
the maximum central temperature @xmath70 and density @xmath71 can be given by @xmath72 and @xmath73 respectively , where @xmath74 and @xmath75 are the central temperature and density of the asteroid in its unperturbed state .
when @xmath76 , the temperature and density may rise enough to detonate effectively a significant fraction of the available thermonuclear fuel , which will affect the orbital evolution of the disrupted debris . in our case , @xmath68 is only about 3 , so we believe that the orbital motion and accretion process will not be significantly modified by this effect . 99 alpar , m. a. , & baykal , a. 2006 , , 372 , 489 anderson , p. w. , & itoh , n. 1975 , , 256 , 25 archibald , r. f. , kaspi , v. m. , ng , c. y. , et al .
2013 , , 497 , 591 backer , d. c. 1973 , , 182 , 245 bilous , a. v. , hessels , j. w. t. , kondratiev , v. i. , et al .
2014 , , 572 , a52 brook , p. r. , karastergiou , a. , buchner , s. , et al . 2014 , , 780 , l31 burgay , m. , possenti , a. , manchester , r. n. , et al .
2005 , , 624 , l113 campana , s. , lodato , g. , davanzo , p. , et al .
2011 , , 480 , 69 carter , b. , & luminet , j. p. 1982
, , 296 , 211 chamel , n. 2013 , physical review letters , 110 , 011101 cognard , i. , & backer , d. c. 2004 , , 612 , l125 colgate , s. a. , & petschek , a. g. 1981 , , 248 , 771 deich , w. t. s. , cordes , j. m. , hankins , t. h. , & rankin , j. m. 1986 , , 300 , 540 dib , r. , kaspi , v. m. , & gavriil , f. p. 2009
, , 702 , 614 duncan , r. c. , & thompson , c. 1992 , , 392 , l9 esamdin , a. , lyne , a. g. , graham - smith , f. , et al .
2005 , , 356 , 59 espinoza , c. m. , lyne , a. g. , stappers , b. w. , & kramer , m. 2011 , , 414 , 1679 gavriil , f. p. , dib , r. , & kaspi , v. m. 2011 , , 736 , 138 glampedakis , k. , & andersson , n. 2009 , physical review letters , 102 , 141101 haskell , b. , pizzochero , p. m. , & sidery , t. 2012 , , 420 , 658 hills , j. g. 1975 , , 254 , 295 hooker , j. , newton , w. g. , & li , b. a. 2013 , arxiv:1308.0031 huang , y. f. , & geng , j. j. 2014 , , 782 , l20 jones , p. b. 2011 , , 414 , 759 jones , p. b. 2014 , , 437 , 4027 karastergiou , a. , roberts , s. j. , johnston , s. , et al . 2011 , , 415 , 251 kaspi , v. m. , lackey , j. r. , & chakrabarty , d. 2000 , , 537 , l31 katz , j. i. 2014 , , 349 , 611 li , j. , esamdin , a. , manchester , r. n. , qian , m. f. , & niu , h. b. 2012 , , 425 , 1294 livingstone , m. a. , kaspi , v. m. , & gavriil , f. p. 2010
, , 710 , 1710 livio , m. , & taam , r. e. 1987 , , 327 , 398 lu , y. , cheng , k. s. , & huang , y. f. 2006 , , 641 , 288 lyutikov , m. 2013 , arxiv:1306.2264 manchester , r. n. , & hobbs , g. 2011 , , 736 , l31 mandal , r. d. r. , konar , s. , dey , m. , & dey , j. 2009 , , 399 , 822 morley , p. d. , & schmidt , i. 1996 , epl ( europhysics letters ) , 33 , 105 olausen , s. a. , & kaspi , v. m. 2014 , , 212 , 6 ouyed , r. , leahy , d. , & koning , n. 2014 , , 352 , 715 page , c. g. 1973 , , 163 , 29 pines , d. , & alpar , m. a. 1985 , , 316 , 27 pizzochero , p. m. 2011 , , 743 , l20 poutanen , j. , ibragimov , a. , & annala , m. 2009 , , 706 , l129 rankin , j. m. 1986 , , 301 , 901 rankin , j. m. , & wright , g. a. e. 2008 , 40 years of pulsars : millisecond pulsars , magnetars and more , 983 , 91 rees , m. j. 1988 , , 333 , 523 ruderman , m. 1976 , , 203 , 206 amaz mu , s. , aydin , b. , gouml , e. 2014 , , 440 , 2916 seveso , s. , pizzochero , p. m. , & haskell , b. 2012 , , 427 , 1089 shakura , n. i. , & sunyaev , r. a. 1973 , , 24 , 337 shannon , r. m. , cordes , j. m. , metcalfe , t. s. , et al .
2013 , , 766 , 5 thompson , c. , duncan , r. c. , woods , p. m. , et al .
2000 , , 543 , 340 thorsett , s. e. , arzoumanian , z. , camilo , f. , & lyne , a. g. 1999 , , 523 , 763 ulmer , a. 1999 , , 514 , 180 usov , v. v. 1992 , , 357 , 472 van buren , d. 1981 , , 249 , 297 wang , n. , manchester , r. n. , pace , r. t. , et al .
2000 , , 317 , 843 wang , z. , chakrabarty , d. , & kaplan , d. l. 2006 , , 440 , 772 warszawski , l. , & melatos , a. 2011 , , 415 , 1611 warszawski , l. , melatos , a. , & berloff , n. g. 2012 , , 85 , 104503 weltevrede , p. , johnston , s. , manchester , r. n. , et al .
2010 , , 27 , 64 wolszczan , a. 1994 , science , 264 , 538 yu , m. , manchester , r. n. , hobbs , g. , et al . 2013 , , 429 , 688 yuan , j. p. , manchester , r. n. , wang , n. , et al .
2010 , , 719 , l111 zhang , b. , xu , r. x. , & qiao , g. j. 2000 , , 545 , l127 zhou , e. p. , lu , j. g. , tong , h. , & xu , r. x. 2014 , , 443 , 2705 | long term observations by brook et al .
reveal that the derivative of rotational frequency of psr j0738 - 4042 changed abruptly in 2005 .
originally , the spin - down rate was relatively stable , with the rotational frequency derivative of @xmath0 .
after september 2005 , the derivative began to rise up .
about 1000 days later , it arrived at another relatively stable value of about @xmath1 , indicating that the pulsar is spinning - down relatively slowly . to explain the observed spin - down rate change , we resort to an asteroid disrupted by psr j0738 - 4042 . in our model , the orbital angular momentum of the asteroid
is assumed to be parallel to that of the rotating pulsar , so that the pronounced reduction in the spin - down rate can be naturally explained as due to the transfer of the angular momentum from the disrupted material to the central pulsar .
the derived magnetospheric radius is about @xmath2 cm , which is smaller than the tidal disruption radius ( @xmath3 cm ) .
our model is self - consistent .
it is shown that the variability of the spin - down rate of psr j0738 - 4042 can be quantitatively accounted for by the accretion from the asteroid disrupted by the central pulsar . |
the abjm theory @xcite provides a rigorous formulation to study the m2-brane dynamics . at the same time
, it also provides a very profound mathematical structure . in studying the partition function or one - point functions of the bps wilson loop ,
many interesting properties were found .
most of these properties are shared with the superconformal chern - simons theories with a large number of supersymmetries @xmath2 or @xmath3 @xcite .
one interesting discovery is the hidden supergroup structure @xcite . after using the localization techniques @xcite ,
the infinite - dimensional path integral in defining the partition function or the one - point functions of the theories is reduced to a finite - dimensional multiple integration .
these matrix models take the form of the gaussian matrix model with two simultaneous deformations : the supergroup deformation and the trigonometric ( or hyperbolic ) deformation . for the abjm theory
we replace the u@xmath4 gauge group of the gaussian matrix model by the supergroup u@xmath5 and at the same time we change the rational functions by hyperbolic functions . in terms of the hidden supergroups ,
other @xmath2 or @xmath3 superconformal theories are associated with u@xmath6 or osp@xmath7 .
another discovery is the fermi gas formalism , which was first proposed in @xcite for the original u@xmath5 abjm theory .
it was found that the partition function of the u@xmath8 theory can also be rewritten into that of a fermi gas system @xcite in the u@xmath8 generalization , see @xcite . ]
@xmath9 where the density matrix @xmath10 relates a state @xmath11 to its permutation @xmath12 which is accompanied by a sign factor @xmath13 . using the position operator @xmath14 corresponding to the state @xmath15 and the dual momentum operator @xmath16 obeying the canonical commutation relation @xmath17=i\hbar$ ] with @xmath18 , the density matrix is explicitly given by @xmath19 with @xcite
@xmath20 the fermi gas formalism is not only beautiful but also practical .
in fact we can follow the systematic wkb ( small @xmath21 ) analysis @xcite to obtain the large @xmath22 expansion of the partition function .
finally , the full large @xmath22 expansion was obtained @xcite , based on the analysis in the fermi gas formalism ( small @xmath21 expansion @xcite and exact values for finite @xmath21 @xcite ) together with the results from the t hooft expansion @xcite .
see @xcite for a review .
after establishing the result for the unitary supergroup , it is interesting to ask what happens if we replace the unitary supergroup by an orthosymplectic supergroup , whose physical interpretation is the introduction of an orientifold plane in the type iib setup .
it was pointed out in @xcite that generally in studying the theories with orthosymplectic groups the projected density matrices introduced in @xcite @xmath23_\pm=\sqrt{v_m(\widehat q ) } \frac{\widehat\pi_\pm}{2\cosh\frac{\widehat p}{2 } } \sqrt{v_m(\widehat q)},\quad \widehat\pi_\pm=\frac{1\pm \widehat r}{2 } , \label{denmatvmproj}\end{aligned}\ ] ] play crucial roles , where @xmath24 is a reflection operator , @xmath25 .
owing to the large number of the supersymmetries , we hope that the non - perturbative effect of the orientifold plane can be clearly identified by studying the theories with orthosymplectic supergroups . with this expectation , recently in @xcite the full large @xmath22 expansion of the osp@xmath26 theory was studied with the even projected density matrix and a relation to the original abjm u@xmath5 theory was found by doubling the orthosymplectic quiver in the sense of @xcite . along these directions , interestingly it was found @xcite that the density matrix for the osp@xmath27 theory is identical to that for the abjm u@xmath5 theory with the odd projection @xmath28_-,\end{aligned}\ ] ] which allows us to study the osp@xmath27 theory directly from the abjm theory .
subsequently this claim was generalized to the case of non - equal ranks @xcite and u@xmath29 theories are complex conjugate to each other , while the density matrices are identical .
though , as a convention , we typically consider @xmath30 with @xmath31 , the schematic pattern may be clearer if we align the ranks of the unitary supergroups with those of the orthosymplectic supergroups . ]
\label{msresult}\end{aligned}\ ] ] related to these results , it was also observed in @xcite that the values of the partition function for the osp@xmath26 theory @xcite match with those for the u@xmath33 theory with the even projection for various integral values of @xmath21 . from this
we naturally expect the relation @xmath34_+ , \label{2n|2n}\end{aligned}\ ] ] though the proof has not been known . in the first part of this paper
, we shall fill this gap .
namely , we first prove by generalizing it to @xmath35_+ , \label{even}\end{aligned}\ ] ] with a non - negative integer @xmath36 .
it is natural to ask what happens to the case when the rank of the orthogonal bosonic subgroup is greater than that of the symplectic subgroup .
we answer this question by proving @xmath37 @xmath38_- .
\label{odd}\end{aligned}\ ] ] the result is schematically summarized in figure [ osppic ] .
combined with the results from @xcite , we find an interesting pattern depicted in table [ relation ] .
all of these relations completely reduce the study of the theories with orthosymplectic supergroups into that for the unitary supergroups with chiral projections .
@xmath39_-\\ \widehat\rho_{\text{osp}(2n|2n)}&=[\widehat\rho_{\text{u}(n|n+1)}]_+ & \widehat\rho_{\text{osp}(2n+2|2n)}&=[\widehat\rho_{\text{u}(n+1|n)}]_-\\ \widehat\rho_{\text{osp}(2n+1|2n+2)}&=[\widehat\rho_{\text{u}(n|n+2)}]_- & \widehat\rho_{\text{osp}(2n+3|2n)}&=[\widehat\rho_{\text{u}(n+2|n)}]_-\\ \widehat\rho_{\text{osp}(2n|2n+2)}&=[\widehat\rho_{\text{u}(n|n+3)}]_+ & \widehat\rho_{\text{osp}(2n+4|2n)}&=[\widehat\rho_{\text{u}(n+3|n)}]_-\\ & \vdots&&\vdots\\ \widehat\rho_{\text{osp}(2n|2n+k-2 ) } & = [ \widehat\rho_{\text{u}(n|n+k-1)}]_+ & \widehat\rho_{\text{osp}(2n+k|2n ) } & = [ \widehat\rho_{\text{u}(n+k-1|n)}]_-\\ \widehat\rho_{\text{osp}(2n+1|2n+k ) } & = [ \widehat\rho_{\text{u}(n|n+k)}]_- & \widehat\rho_{\text{osp}(2n+k+1|2n ) } & = [ \widehat\rho_{\text{u}(n+k|n)}]_-\end{aligned}\ ] ] the key observation in our proof is to utilize the extra hyperbolic sine function @xmath40 appearing in the measure of the bosonic symplectic subgroup .
although this factor was cumbersome in the fermi gas formalism proposed in @xcite , owing to this factor we can naturally introduce a hyperbolic tangent function in the matrix elements , which is fourier - dual to a hyperbolic cosecant function @xcite , reproducing the same factor in for odd @xmath41 .
in @xcite only the fermi gas system for the density matrices with the odd projection is assigned a physical meaning from the theories with orthosymplectic supergroups .
here we have seen that both the odd and even chiral projections are physically relevant .
note that the relations and are consistent with the duality .
the duality for the theories with orthosymplectic supergroups @xcite for the orthosymplectic theories is different from that in @xcite : @xmath42 . ]
@xmath43 is translated to @xmath44 which is consistent with the duality for the theories with unitary supergroups . in @xcite
the duality leads respectively to the constraint @xmath45 for the osp@xmath46 theories and @xmath47 for the osp@xmath48 theories . for the osp@xmath48 theories
was proposed in @xcite , in our analysis we find it more natural to exclude the @xmath49 case and the dual @xmath50 case in the osp@xmath48 theories . ] after establishing the relation between the theories with orthosymplectic supergroups and those with unitary supergroups with the chiral projections , in the second part we proceed to study the instanton effects @xcite in the chirally projected @xmath51 theories .
first of all we introduce the chemical potential @xmath52 dual to the particle number @xmath22 and switch from the fermi gas partition function to the grand potential @xmath53 defined by @xcite @xmath54 where on the right - hand side we have combined the permutations among the @xmath22 integration variables in into the fredholm determinant .
as its periodicity in @xmath55 causes an oscillating behavior , it is reasonable to decompose the grand potential into the non - oscillating part @xmath56 and the oscillations as @xmath57.\end{aligned}\ ] ] hereafter the original grand potential @xmath53 is referred to as the _ full _ grand potential while the non - oscillating part @xmath56 as the _ modified _ grand potential .
the instanton effects appear as the non - perturbative effects in the chemical potential @xmath58 in the large @xmath52 expansion of the modified grand potential @xmath56 . to investigate the instanton effects in the chirally projected theories
, we shall consider the two combinations @xmath59 and @xmath60 , instead of @xmath61 , the original modified grand potentials for the even and odd projected systems . in the previous works @xcite the modified grand potentials
are studied for @xmath62 and @xmath63 , and it was observed that there are two kinds of the instanton effects , the worldsheet instantons @xmath64 and the membrane instantons @xmath65 @xmath66 .
especially it was observed that the anti - symmetric combination @xmath67 contains only the membrane instantons @xmath68 , which was also studied in full detail by the wkb expansion in @xcite .
our main interest in this paper is the instantons in the symmetric combination @xmath69 .
this part contains not only the membrane instantons but also the worldsheet instantons @xmath70 , which are non - perturbative in @xmath21 and thus can not be analyzed in the wkb expansion .. it was observed in @xcite , however , that these effects are completely absorbed by the shift of the chemical potential @xmath52 into the effective chemical potential @xmath71 which is defined in the same way as in the abjm theory without chiral projections @xcite . ] instead , for this part we can utilize a trivial relation among the full grand potential for @xmath72 and those for @xmath73_\pm$ ] @xmath74 which is , in turn , translated to a relation among @xmath69 , @xmath67 and the modified grand potential of the unprojected system @xmath56 .
note that this relation also implies that the perturbative part and the membrane instantons in @xmath69 completely coincide with those in the system without chiral projections .
in this paper we study this relation carefully . by taking the explicit results for @xmath67 into account
, we finally identify a simple relation without the oscillatory behavior , @xmath75 the result is derived in a similar way as for the duplicate quivers @xcite .
with this relation the worldsheet instanton in @xmath69 is completely solved in terms of the result of the abjm theory .
surprisingly , we find that the worldsheet instanton in the chirally projected theories fits well with the gopakumar - vafa formula for the topological string partition function @xcite .
for the worldsheet instanton in the unprojected @xmath51 theories , it was already known that the same formula with the topological invariants on local @xmath76 works @xcite .
it is still non - trivial that the formula applies for the chirally projected theories since the relation induces new instanton effects non - linearly . nevertheless , with we
can check it and identify the gopakumar - vafa invariants @xmath77 for the chirally projected theories up to the seventh instanton as in tables [ gv12345 ] and [ gv67 ] . at present the interpretation of the invariants
@xmath77 is unclear .
we shall briefly argue this point in section [ discussion ] .
the organization of this paper is as follows . in section [ chiral ]
, we prove the relations and . after reducing the question for the theories with orthosymplectic supergroups into that for the theories with unitary supergroups and the chiral projections , in section [ ws ] we study the grand potential for the latter theories .
finally we conclude with discussions in section [ discussion ] .
in this section , we prove that the fermi gas system for the osp@xmath46 theory with @xmath45 matches with that for the u@xmath78 theory with the even projection , while the fermi gas system for the osp@xmath48 theory with @xmath47 is that for the u@xmath79 theory with the odd projection .
namely , we study the partition function of the theory with the orthosymplectic supergroup carefully and finally find that the partition function is rewritten into @xmath80 with some function @xmath81 and @xmath82 given by @xmath83 the right - hand side of is known to be written as the partition function of the @xmath22 particle ideal fermi gas system with the density matrix @xmath84 which coincide with the chirally projected density matrix for the @xmath85 theory if @xmath86 . the proof goes mostly in parallel with @xcite except only a few important novelties .
in this subsection we provide the proof of . after the localization technique ,
the partition function is @xcite @xmath87 with the one - loop determinant factor @xmath88 as in @xcite our starting point is the cauchy - vandermonde determinant @xmath89_{(i , k)\in z_n\times z_{n+m}}\\ \bigl[\frac{\sinh\frac{m\nu_k}{k}}{\sinh\frac{\nu_k}{k}}\bigr ] _ { ( m , k)\in z_m\times z_{n+m } } \end{pmatrix}=(-1)^{mn+\frac{1}{2}m(m-1)}\nonumber\\ & \quad\times\frac{\prod_{i < j}^n(2\sinh\frac{\mu_i-\mu_j}{2k } ) ( 2\sinh\frac{\mu_i+\mu_j}{2k } ) \prod_{k < l}^{n+m}(2\sinh\frac{\nu_k-\nu_l}{2k } ) ( 2\sinh\frac{\nu_k+\nu_l}{2k } ) } { \prod_{i=1}^m\prod_{k=1}^{n+m}(2\cosh\frac{\mu_i-\nu_k}{2k } ) ( 2\cosh\frac{\mu_i+\nu_k}{2k } ) } , \label{cauchyvdm}\end{aligned}\ ] ] where we have denoted the integer set spanned by each index with the notation @xmath90 . after multiplying @xmath91 to each column , we obtain the following determinant expression for @xmath92 @xmath93 _ { ( i , k)\in z_n\times z_{n+m}}\\ \bigl[2\sinh\frac{m\nu_k}{k}\bigr]_{(m , k)\in z_m\times z_{n+m } } \end{pmatrix}^2 . \label{1-loop_in_det}\end{aligned}\ ] ] to proceed , it is helpful to introduce the canonical position and momentum operators @xmath14 and @xmath16 obeying the commutation relation @xmath17=i\hbar$ ] with @xmath18 .
we shall also introduce the states @xmath94 to abbreviate the matrix components in the lower block of @xmath95}\end{aligned}\ ] ] where @xmath96 is the coordinate eigenstate normalized as @xmath97 .
for the upper block , it is crucial to express the matrix component as is cancelled by the projections and hence harmless . ]
@xmath98 using the fourier duality between hyperbolic tangent functions and hyperbolic cosecant functions @xcite .
this is the most important point in our proof . after noticing this structure
the rest of the computation is quite parallel to @xcite : substitute @xmath92 with and ; include the fresnel factors @xmath99 and @xmath100 into the brackets ; trivialize the first determinant by renaming the indices of @xmath101 ; perform the similarity transformation @xmath102 to the integration variables . after these steps we obtain the expression @xmath103 _ { ( k , j)\in z_{n+m}\times z_n } & \bigl[\langle\nu_k|e^{\frac{i}{2\hbar}\widehat p^2}|n\rrb\bigr ] _
{ ( k , n)\in z_{n+m}\times z_m}\end{pmatrix}. \label{zeven2}\end{aligned}\ ] ] as in @xcite the elements in the two products in front of the determinant become the delta functions and we can perform the integrations .
we shall see this explicitly in the following .
let us first simplify the determinant .
using @xmath104 the determinant reduces to @xmath105_{z_{n+m}\times z_n } & \bigl [ \langle\nu_k|e^{\frac{i}{2\hbar}\widehat p^2}|n\rrb \bigr]_{z_{n+m}\times z_m } \end{pmatrix}\nonumber\\ & = e^{-\frac{i}{12\hbar}(2\pi)^2m(2m+1)(m+1 ) } \det\begin{pmatrix } \bigl [ -2ik\langle\nu_k|\frac{1}{2\sinh\frac{\widehat p}{2 } } \widehat\pi_+|\mu_j\rangle \bigr]_{z_{n+m}\times z_n } & \bigl [ \langle\nu_k|n\rrb \bigr]_{z_{n+m}\times z_m } \end{pmatrix } , \label{evendet}\end{aligned}\ ] ] which is an odd function of @xmath101 .
the two products in the second line of can be formally computed as @xmath106 \sqcup ( 2\pi im,\infty+2\pi im)\end{aligned}\ ] ] so that the new contour contains the supports of the delta functions .
we can show that such a deformation is allowed if and only if @xmath107 , following the argument in @xcite . ]
@xmath108 since the rest of the integrand is an odd function of @xmath101 , we can drop one of the two delta functions as in @xcite , @xmath109 after reducing the two matrix elements into the delta functions we can perform the @xmath101 integrations by simple substitutions , leaving only a single determinant in . reverting it into the products and separating the @xmath22-independent factors
, we finally find that the partition function is given by @xmath110 as @xmath111 with @xmath112 being an @xmath22-independent factor @xmath113 ( which is non - vanishing for @xmath45 ) and @xmath114 given by @xmath115 with @xmath116 .
comparing with the expression we find that @xmath117 . in this way we have proved that the fermi gas system for the osp@xmath46 theory is identical to that for the u@xmath78 theory with the even chiral projection . in the previous subsection
, we have seen that the density matrix for the theory with orthosymplectic supergroups where the rank of the bosonic symplectic subgroup is greater than or equal to that of the orthogonal subgroup is related to that for the unitary supergroups . here
we shall turn to the opposite case : osp@xmath48 with @xmath47 .
the partition function is @xcite @xmath118 since we can add one row by a multiple of another row in the determinant without changing its value , we can express the cauchy - vandermonde determinant by @xmath89_{(i , k)\in z_{n+m}\times z_{n } } & \bigl[2\cosh\frac{(m-1)\mu_i}{k}\bigr ] _ { ( m , k)\in z_{n+m}\times z_{m } } \end{pmatrix}\nonumber\\ & = 2(-1)^{mn+\frac{1}{2}m(m-1 ) } \frac{\prod_{i < j}^{n+m}(2\sinh\frac{\mu_i-\mu_j}{2k } ) ( 2\sinh\frac{\mu_i+\mu_j}{2k } ) \prod_{k < l}^{n}(2\sinh\frac{\nu_k-\nu_l}{2k } ) ( 2\sinh\frac{\nu_k+\nu_l}{2k } ) } { \prod_{i=1}^{n+m}\prod_{k=1}^{n}(2\cosh\frac{\mu_i-\nu_k}{2k } ) ( 2\cosh\frac{\mu_i+\nu_k}{2k})}.\end{aligned}\ ] ] after multiplying @xmath91 to each column , the left block in the determinant can be expressed as the previous case .
for the right block we introduce the states @xmath119 by @xmath120 then , through the same steps as in the last subsection , we can rewrite the partition function as @xmath121_{(k , j)\in z_{n}\times z_{n+m}}\\ \bigl[\llp n|e^{-\frac{i}{2\hbar}\widehat p^2}|\mu_j\rangle\bigr ] _ { ( n , j)\in z_{m}\times z_{n+m } } \end{pmatrix},\end{aligned}\ ] ] where the first two factors can be replaced with the delta functions @xmath122 after performing the integration over @xmath123 , this time we obtain @xmath124 with @xmath125 and @xmath126 hence again is nothing but the fermi gas formalism for the @xmath127 theory with the odd chiral projection .
in the previous section , combined with the result from @xcite , we have found that the matrix model associated with any orthosymplectic supergroup is directly related to that associated with the corresponding unitary supergroup and the chiral projection .
this means that we can study the osp@xmath7 theories with completely general @xmath128 from the @xmath85 theories with the chiral projections .
here we proceed to study the grand potentials for the latter theories .
first we define the following symmetric and anti - symmetric combination @xmath129 where @xmath61 are respectively the modified grand potential for the density matrix @xmath73_\pm$ ] .
the large @xmath52 expansion of @xmath69 and @xmath67 was studied previously in @xcite , and was observed to have the following structures @xmath130 with @xmath131 and @xmath132 being some constants depending on @xmath21 and @xmath41 . here @xmath71 is the effective chemical potential defined in the @xmath85 theory without chiral projections @xcite @xmath133 with some constants @xmath134 .
the possible bound states of the worldsheet instantons and the membrane instantons @xmath135 are observed to be absorbed into the worldsheet instantons @xmath136 through the non - perturbative deviations in @xmath71 .
note that @xmath69 and the modified grand potential without chiral projections @xmath56 have the same schematic structure : they have completely the same perturbative part and the membrane instantons ; they are different only in the coefficients of the worldsheet instantons .
several coefficients of the half membrane instantons @xmath137 @xmath138 in @xmath67 are determined as explicit functions of @xmath139 through the extrapolation of the wkb small @xmath21 expansion in @xcite . below we shall study the worldsheet instantons in the symmetric combination @xmath69 . with the help of the schematic expression of the anti - symmetric part @xmath67
, we extract from a new relation between @xmath69 and @xmath56 . using it we further compute the explicit expression of the first few instanton coefficients and find that they fit with the gopakumar - vafa formula of the topological string theory , with the gopakumar - vafa invariants identified as in tables [ diaggv ] , [ gv12345 ] and [ gv67 ] . from the fundamental properties of the projection operators
@xmath140 and @xmath141 , it follows that @xmath142_+)\det(1+e^\mu[{\widehat \rho}_{\text{u}(n|n+m)}]_-).\end{aligned}\ ] ] this is translated into the relations among the full grand potentials or @xmath143 at first sight it seems difficult to extract a relation between the modified grand potentials as the equation contains the oscillating terms both for @xmath56 and @xmath61 .
as we see below , however , using the schematic expression of the anti - symmetric part @xmath67 we obtain a more refined relation where the oscillations for @xmath56 are absent .
we rewrite the relation in terms of @xmath69 and @xmath67 @xmath144 as @xmath67 in has the following simple quasi - periodicity @xmath145 the relation reduces to @xmath146 here we notice that the terms with odd @xmath147 do not contribute , since each of those terms is always accompanied with the one with the opposite sign @xmath148 then , can be rewritten as @xmath149 this relation can be solved by @xmath150 which can be rewritten as .
as in @xcite we can easily check that correctly reproduces and contains no oscillations . to obtain the direct relation between the worldsheet instantons in @xmath69 and those in @xmath56 ,
let us compute the exponent inside the logarithm in using the schematic expression of @xmath69 .
note that , since the effective chemical potential @xmath71 is related to the original one @xmath52 by , shifting @xmath52 by @xmath151 directly means shifting @xmath71 .
hence , in the following we shall regard @xmath69 in as a function of @xmath71 and , with a slight abuse of notation , denote the same function as @xmath152 .
then , most of the contributions in the exponent in including the non - perturbative membrane instantons cancel and we are left with the @xmath153 term and the worldsheet instantons as @xmath154 where we have used the explicit value of @xmath153 . hence we obtain an equation containing only the worldsheet instantons @xmath155 where @xmath156 and @xmath157 is the worldsheet instantons in the @xmath51 theory without chiral projections @xcite @xmath158 we can solve inversely order by order in @xmath159 as @xmath160 then , we find that the functional form of @xmath161 reproduces all the numerical fitting in @xcite .
more surprisingly , this functional form fit well with the gopakumar - vafa formula , which is not guaranteed from the beginning , @xmath162 where we have written the arguments @xmath139 to express the multi - covering structure of the coefficients .
we have also introduced the extra factor of @xmath163 in front of the gopakumar - vafa invariants @xmath164 , as the original modified grand potentials @xmath61 associated to the density matrices @xmath165_\pm$ ] are related to @xmath69 and @xmath67 by @xmath166 .
.the diagonal gopakumar - vafa invariants @xmath167 @xmath168 identified for the chirally projected theory @xmath169 . [ cols="^,^,^,^,^,^,^,^",options="header " , ]
in this paper we have proved that the density matrix of the theories with orthosymplectic supergroups reduces to that of the theories with unitary supergroups and the chiral projections .
we have further established the identity between the total grand potential @xmath56 and the sum of the projected ones @xmath69 . from this identity
we can derive the gopakumar - vafa invariants systematically without mentioning to the numerical fitting , as long as we know those for the theories with unitary supergroups .
let us raise several questions related to our results and discuss some further directions .
firstly , though we have found the equivalences between the chiral projections of the @xmath85 theories and the theories with the orthosymplectic groups , the counterpart of the u@xmath170 theory is still missing ( see table [ relation ] ) .
it is then interesting to ask the role of this sector in the context of the orientifold background .
secondly , we shall argue possible extension of our computations .
we have introduced the states @xmath94 , @xmath119 in section [ chiral ] and @xmath171 in @xcite .
it is interesting to note that these combinations appear naturally in the weyl character formulas for sp@xmath172 , o@xmath172 and o@xmath173 respectively ( see e.g. @xcite ) .
apparently , this interpretation is a good sign for generalizations of our results on partition functions into one - point functions of the bps wilson loops @xcite .
it is also interesting to study the non - perturbative effects of the orientifold plane for other exactly solvable chern - simons theories like the @xmath174 model @xcite .
thirdly , we have found that the worldsheet instanton in the chirally projected theories fits with the gopakumar - vafa formula , though the interpretation of the gopakumar - vafa invariants in tables [ diaggv ] , [ gv12345 ] and [ gv67 ] is unclear to us . from our knowledge of the unprojected theories ,
it is natural to expect a relation to local @xmath175 with a projection
. it will be interesting to understand the gopakumar - vafa invariants from the geometrical viewpoint .
finally let us discuss the bps index .
in general the worldsheet instantons are changed drastically , while the membrane instantons remain unmodified .
this indicates that , if we assume the same refined topological string expression as in the abjm theory @xcite , the bps index is reshuffled with the membrane instantons kept fixed .
for example , for @xmath176 , the only non - vanishing bps index in the abjm theory is @xmath177 . in the orientifold theory , if we make an ansatz that only the bps indices for @xmath178 and @xmath179 are non - vanishing from the expression of the instantons , we find non - trivial relations @xmath180 by matching the expressions of the instantons
. then , apparently there are no non - negative solutions .
we would like to search for the correct analysis for the topological invariants and see the reshuffling more clearly in future .
geometry to that of local @xmath181 when moving the ellipsoid deformation parameter from @xmath182 to @xmath183 .
the change of the bps indices may be similar to our situation . ]
we are grateful to masazumi honda , kazumi okuyama and takao suyama for valuable discussions .
the work of s.m . is supported by jsps grant - in - aid for scientific research ( c ) # 26400245 .
99 o. aharony , o. bergman , d. l. jafferis and j. maldacena , `` n=6 superconformal chern - simons - matter theories , m2-branes and their gravity duals , '' jhep * 0810 * , 091 ( 2008 ) [ arxiv:0806.1218 [ hep - th ] ] .
k. hosomichi , k. m. lee , s. lee , s. lee and j. park , `` n=5,6 superconformal chern - simons theories and m2-branes on orbifolds , '' jhep * 0809 * , 002 ( 2008 ) [ arxiv:0806.4977 [ hep - th ] ] .
o. aharony , o. bergman and d. l. jafferis , `` fractional m2-branes , '' jhep
* 0811 * , 043 ( 2008 ) [ arxiv:0807.4924 [ hep - th ] ] . n. drukker and d. trancanelli , `` a supermatrix model for n=6 super chern - simons - matter theory , '' jhep * 1002 * , 058 ( 2010 ) [ arxiv:0912.3006 [ hep - th ] ] .
m. marino and p. putrov , `` exact results in abjm theory from topological strings , '' jhep * 1006 * , 011 ( 2010 ) [ arxiv:0912.3074 [ hep - th ] ] .
a. kapustin , b. willett and i. yaakov , `` exact results for wilson loops in superconformal chern - simons theories with matter , '' jhep * 1003 * , 089 ( 2010 ) [ arxiv:0909.4559 [ hep - th ] ] .
m. marino and p. putrov , `` abjm theory as a fermi gas , '' j. stat .
mech . * 1203 * , p03001 ( 2012 ) [ arxiv:1110.4066 [ hep - th ] ]
. h. awata , s. hirano and m. shigemori , `` the partition function of abj theory , '' ptep * 2013 * , 053b04 ( 2013 ) [ arxiv:1212.2966 ] . m. honda , `` direct derivation of `` mirror '' abj partition function , '' jhep * 1312 * , 046 ( 2013 ) [ arxiv:1310.3126 [ hep - th ] ] . s. matsumoto and s. moriyama , `` abj fractional brane from abjm wilson loop , '' jhep * 1403 * , 079 ( 2014 ) [ arxiv:1310.8051 [ hep - th ] ] . m. honda and k. okuyama , `` exact results on abj theory and the refined topological string , '' jhep * 1408 * , 148 ( 2014 ) [ arxiv:1405.3653 [ hep - th ] ] . y. hatsuda , m. marino , s. moriyama and k. okuyama , `` non - perturbative effects and the refined topological string , '' jhep * 1409 * , 168 ( 2014 ) [ arxiv:1306.1734 [ hep - th ] ] . f. calvo and m. marino , `` membrane instantons from a semiclassical tba , '' jhep * 1305 * , 006 ( 2013 ) [ arxiv:1212.5118 [ hep - th ] ] . y. hatsuda , s. moriyama and k. okuyama , `` exact results on the abjm fermi gas , '' jhep * 1210 * , 020 ( 2012 ) [ arxiv:1207.4283 [ hep - th ] ] .
p. putrov and m. yamazaki , `` exact abjm partition function from tba , '' mod .
a * 27 * ( 2012 ) 1250200 [ arxiv:1207.5066 [ hep - th ] ] .
y. hatsuda , s. moriyama and k. okuyama , `` instanton effects in abjm theory from fermi gas approach , '' jhep * 1301 * , 158 ( 2013 ) [ arxiv:1211.1251 [ hep - th ] ]
. y. hatsuda , s. moriyama and k. okuyama , `` instanton bound states in abjm theory , '' jhep * 1305 * , 054 ( 2013 ) [ arxiv:1301.5184 [ hep - th ] ] .
n. drukker , m. marino and p. putrov , `` from weak to strong coupling in abjm theory , '' commun .
phys . * 306 * , 511 ( 2011 ) [ arxiv:1007.3837 [ hep - th ] ] .
n. drukker , m. marino and p. putrov , `` nonperturbative aspects of abjm theory , '' jhep * 1111 * , 141 ( 2011 ) [ arxiv:1103.4844 [ hep - th ] ] .
h. fuji , s. hirano and s. moriyama , `` summing up all genus free energy of abjm matrix model , '' jhep * 1108 * , 001 ( 2011 ) [ arxiv:1106.4631 [ hep - th ] ] .
m. hanada , m. honda , y. honma , j. nishimura , s. shiba and y. yoshida , `` numerical studies of the abjm theory for arbitrary n at arbitrary coupling constant , '' jhep * 1205 * , 121 ( 2012 ) [ arxiv:1202.5300 [ hep - th ] ] .
y. hatsuda , s. moriyama and k. okuyama , `` exact instanton expansion of the abjm partition function , '' ptep * 2015 * , no .
11 , 11b104 ( 2015 ) [ arxiv:1507.01678 [ hep - th ] ] . m. mezei and s. s. pufu , `` three - sphere free energy for classical gauge groups , '' jhep * 1402 * , 037 ( 2014 ) [ arxiv:1312.0920 [ hep - th ] ] . k. okuyama , `` probing non - perturbative effects in m - theory on orientifolds , '' jhep * 1601 * , 054 ( 2016 ) [ arxiv:1511.02635 [ hep - th ] ] . s. moriyama and t. suyama , `` instanton effects in orientifold abjm theory , '' arxiv:1511.01660 [ hep - th ]
. m. honda and s. moriyama , `` instanton effects in orbifold abjm theory , '' jhep * 1408 * , 091 ( 2014 ) [ arxiv:1404.0676 [ hep - th ] ] .
m. honda , `` exact relations between m2-brane theories with and without orientifolds , '' arxiv:1512.04335 [ hep - th ] .
s. moriyama and t. suyama , `` orthosymplectic chern - simons matrix model and chirality projection , '' arxiv:1601.03846 [ hep - th ] .
k. okuyama , `` orientifolding of the abj fermi gas , '' arxiv:1601.03215 [ hep - th ] .
b. assel , n. drukker and j. felix , `` partition functions of 3d @xmath184-quivers and their mirror duals from 1d free fermions , '' jhep * 1508 * , 071 ( 2015 ) [ arxiv:1504.07636 [ hep - th ] ] . s. moriyama and t. nosaka , `` superconformal chern - simons partition functions of affine d - type quiver from fermi gas , '' jhep * 1509 * , 054 ( 2015 ) [ arxiv:1504.07710 [ hep - th ] ] . k. becker , m. becker and a. strominger , `` five - branes , membranes and nonperturbative string theory , '' nucl .
b * 456 * ( 1995 ) 130 [ hep - th/9507158 ] .
r. gopakumar and c. vafa , `` m theory and topological strings .
2 . , '' hep - th/9812127 .
d. r. gulotta , c. p. herzog and t. nishioka , `` the abcdef s of matrix models for supersymmetric chern - simons theories , '' jhep * 1204 * ( 2012 ) 138 [ arxiv:1201.6360 [ hep - th ] ] .
w. fulton and j. harris .
`` representation theory , '' springer , graduate texts in mathematics , 129 ( 1991 ) .
y. hatsuda , m. honda , s. moriyama and k. okuyama , `` abjm wilson loops in arbitrary representations , '' jhep * 1310 * , 168 ( 2013 ) [ arxiv:1306.4297 [ hep - th ] ] . s. moriyama and t. nosaka , `` partition functions of superconformal chern - simons theories from fermi gas approach , '' jhep * 1411 * , 164 ( 2014 ) [ arxiv:1407.4268 [ hep - th ] ] . s. moriyama and t. nosaka , `` exact instanton expansion of superconformal chern - simons theories from topological strings , '' jhep * 1505 * , 022 ( 2015 ) [ arxiv:1412.6243 [ hep - th ] ] .
y. hatsuda , `` abjm on ellipsoid and topological strings , '' arxiv:1601.02728 [ hep - th ] . | we study the partition function of the orientifold abjm theory , which is a superconformal chern - simons theory associated with the orthosymplectic supergroup .
we find that the partition function associated with any orthosymplectic supergroup can be realized as the partition function of a fermi gas system whose density matrix is identical to that associated with the corresponding unitary supergroup with a projection to the even or odd chirality .
furthermore we propose an identity which gives directly all of the gopakumar - vafa invariants for the worldsheet instanton effects in the chirally projected theories .
ocu - phys 442 , yitp-16 - 18 * orientifold abjm matrix model : + chiral projections and worldsheet instantons * + + @xmath0_department of physics , graduate school of science , osaka city university + 3 - 3 - 138 sugimoto , sumiyoshi , osaka 558 - 8585 , japan _ + @xmath1_yukawa institute for theoretical physics ,
kyoto university + kitashirakawa - oiwakecho , sakyo , kyoto 606 - 8502 , japan _ |
electronic states of atoms and molecules are characterized in the lowest level of approximation ( the mean field or hartree - fock description ) in terms of electron occupancy of single particle states corresponding to motion in an average potential .
thus , for example , the ground state of the helium atom is described as @xmath0 @xmath1 viz two electrons in the spin singlet state in the lowest single particle orbital ( in conformity with the pauli exclusion principle ) .
the first excited state is that of the excited configuration expressed in the standard spectroscopic notation as @xmath2 @xmath3 and so on .
various excited states of the helium atom may be excited from the ground state through the scattering of electrons or photons from the helium atom and would show up as peaks ( or breit - wigner resonances ) in the scattering cross - section , as also via the excitation spectrum mapping the inelastic processes .
of course our description through the electronic configuration is only approximate and the true state is better described theoretically by considering the effect of perturbations caused by the interactions beyond the mean field resulting in the mixing of configurations .
a qualitatively different situation arises , however , when we consider the excitation of a sufficiently high lying ( energy wise ) configuration , say for instance the @xmath4 @xmath5 state of the helium atom which possesses an energy above the lowest ionization threshold of helium .
such a state if excited would be auto - ionizing viz .
would decay into @xmath6 .
indeed , these manifest themselves as highly asymmetric peaks in the excitation spectrum .
a qualitative understanding of this phenomenon was provided by rice @xcite and by fano @xcite .
quantitative treatment of the process , rendered somewhat subtle by the fact that it involves the mixing between a discrete state and a continuum ( with an attendant degeneracy in energy ) had to await almost a quarter of a century @xcite .
there has been renewed interest in the fano resonance in recent years in connection with interference effects in the quantum dot @xcite and in the aharonov - bohm ring with a quantum dot embedded in one arm@xcite .
accordingly , it is desirable to expose the essence of the mechanism underlying the fano resonance through exactly soluble models avoiding complications existing in realistic systems inessential for a basic understanding of the highly asymmetric excitation spectrum for a quasi - bound state lying in the continuum . in order to obtain a simple realisation of such a situation
consider the elastic scattering of a particle by a target system describable in terms of a potential @xmath7 .
suppose that the target has an excited state of energy @xmath8 above its ground state .
when the projectile energy @xmath9 is above the threshold @xmath8 then , apart from the elastic scattering the inelastic process is also possible .
one may describe the situation in terms of two channels , the elastic and the inelastic .
the latter channel corresponds to the scattering of the projectile of energy @xmath10 from the excited state of the system through a potential @xmath11 , while the two channels are coupled to each other via a potential @xmath12 . with @xmath13 and @xmath14 denoting the states in the elastic and inelastic channels respectively , the system is governed by the coupled schrdinger equations @xmath15 where @xmath16 is the kinetic energy operator for the projectile .
now if the projectile and excited target system interacting via the potential @xmath11 has a bound state whose energy will naturally be below @xmath8 then we do have a situation where the would - be bound state of the excited channel lies in the continuum of the elastic channel , the desideratum for a fano resonance .
we shall consider three illustrations by choosing the underlying potentials and show that the probability for the excitation of the quasi - bound auto - ionizing state ( so to say ) exhibits a resonance at some energy @xmath17 with a finite width @xmath18 corresponding to a complex pole of the corresponding amplitude of the form @xmath19 similar to what occurs in breit - wigner resonance .
however , that apart the amplitude also possesses a concomitant zero at some real energy @xmath20 near the resonance energy @xmath21 .
it is the juxtaposition of a complex pole and a real zero of the excitation probability in the energy plane which gives rise to the observed highly asymmetric peak as the energy passes through that corresponding to the auto - ionizing level .
the generic pole - zero structure of the amplitude can be encapsulated in the form @xmath22 with @xmath23 measuring the separation of the energy from the resonance position in units of @xmath24 viz . @xmath25 ( the reduced energy variable ) and the asymmetry described by the fano parameter @xmath26 .
whence the excitation line shapes in the neighbourhood of the fano resonance is given by @xmath27 wherein to emphasize the main features the other weakly varying factors have been appropriately factored out
. these line shapes @xcite for different values of the asymmetry parameter @xmath28 are shown in fig .
[ fig : fano1 ] .
the asymmetry parameter @xmath28 may be positive or negative depending on the relative disposition of the pole and the zero viz .
the sign of @xmath29 .
the curves for negative values of @xmath28 may be obtained by changing @xmath30 .
we consider a soluble realisation of the two channel problem with a particle of mass @xmath31 moving in one dimension where we take @xmath32 while @xmath33 and @xmath34 with @xmath35 .
taking @xmath32 is tantamount to saying that scattering in the elastic channel is only due to its coupling to the inelastic channel ; this extreme simplification is to set forth in bold relief the fano mechanism as in such a situation the bound state in the continuum plays a central role since but for the coupling between channels there is no effect ( scattering ) in the elastic channel . defining @xmath36 , @xmath37 , @xmath38 and
@xmath39 the system equations [ vide eq .
( [ eq : sch1 ] ) & eq .
( [ eq : sch2 ] ) ] become @xmath40 we begin by discussing the region above the inelastic threshold ( @xmath41 ) viz we have @xmath42 . since the fano paper @xcite discusses the entire problem using the partial wave phase shifts we also prefer to cast the one dimensional transmission and reflection solutions in terms of eigen - channels @xcite ( analogous to angular momentum
partial waves for central potentials see appendix a ) . noting that our equations of motion eq .
( [ eq : sch1n ] ) & eq .
( [ eq : sch2n ] ) are symmetric under the operation @xmath43 we deduce that our solutions @xmath44 & @xmath45 may be classified into sets that are even or odd under the transformation . with an attractive delta function potential in one dimension
there is always one & only one bound state which is in fact symmetric under parity .
also the odd ( or anti - symmetric ) function vanishes at @xmath46 & hence the delta function potentials are ineffective
. therefore the only nontrivial solutions of our problem are the ones that are even under @xmath43 . we observe that for @xmath47 ( when two channels are uncoupled ) the elastic channel corresponds to a free particle and has only plane wave solutions , while in the inelastic channel we have continuum solutions ( taken to be box - normalised ) described by @xmath48 \label{eq:2prop}\\ \hspace{-5.5cm}\text{\ & \ , a bound state solution}\hspace{1 cm } \phi _ 0 ( x ) & = & \sqrt{\beta _ 0 } e^{-\beta _ 0 |x| } \label{eq:2evans } \end{aligned}\ ] ] @xmath49 we only consider the nontrivial solutions even under @xmath43 .
the bound state occurs at @xmath50 .
if @xmath51 , the bound state in the inelastic channel ( lying of course below its threshold ) is in the region wherein the continuum corresponding to elastic channel occurs .
also the bound state corresponds to a pole at @xmath52 of @xmath53 .
now let us consider the effect of coupling between the two channels .
with particles incident in the elastic channel we must solve eq .
( [ eq : sch1n ] ) & eq .
( [ eq : sch2n ] ) subject to the behaviour @xmath54 \label{eq : soln1}\\ \text{and } \hspace{2 cm } \phi(x ) = \frac{a}{\sqrt{2\pi}}e^{iq|x| } \label{eq : soln2}\end{aligned}\ ] ] since in the inelastic channel we have only outgoing waves .
we obtain @xmath55 and @xmath56 , which follows directly from eq .
( [ wavampl - elas ] ) & eq .
( [ wavampl - inelas ] ) and is a consequence of unitarity with respect to the two channels elastic and inelastic .
the factor of @xmath57 arising from the ratio of the currents in the inelastic and elastic channels and the factor of @xmath58 representing particles in the inelastic channel emerging left - wards and right - wards .
we may also observe that the pole of @xmath59 has shifted as compared to that of @xmath53 as an effect of the coupling embodied by @xmath60 and to find its location we have to solve a quartic equation in @xmath61 .
we may also note that as @xmath62 as it should since for simplicity we have taken @xmath63 . rather than resorting to numerical results let us solve for the pole position ( @xmath64 ) for small @xmath60 .
we find readily ( see appendix b ) that @xmath65 thus for small @xmath60 the pole moves off from the real axis in the @xmath61-plane and sits in the fourth quadrant close to the real axis .
therefore , when we map onto the energy @xmath66-plane this pole vis a vis the unitarity cut ( with branch point at @xmath67 ) lies in the fourth quadrant of the second riemann sheet , as befits a resonance and accordingly @xmath68 - i \,\frac{\hbar^2}{2 m}\ , { \tilde { { \lambda}}_c}^2 \,\frac{{\tilde { { \lambda } } } } { 4({\tilde { { \delta } } } - \frac{{\tilde { { \lambda}}}^2}{4})^{1/2}}\ ] ] for a physical understanding of the width @xmath69 of the resonance we recognize that this ` bound state in a continuum ' or the ` auto - ionizing ' state having an unperturbed energy @xmath70 decays into the continuum states of the elastic channel caused by the interaction @xmath71 . to obtain the rate of the decay for small @xmath72 it would suffice to use time - dependent perturbation theory or the fermi golden rule viz .
@xmath73 where initial state is
the unperturbed bound state described by the wave - function @xmath74 and the final unperturbed state is @xmath75 ( taking box normalisation say ) and the number of states of a free particle between momenta @xmath76 and @xmath77 is @xmath78 , which gives @xmath79 .
accordingly , the decay rate for the ` auto - ionizing ' state to go into the continuum is given by @xmath80 in agreement with our earlier result describing the excursion of the pole into the complex plane .
further one may use perturbation theory to verify that the shift in the real part of energy to order @xmath81 is zero . in order to understand the origin of the fano zero
we must examine the situation at energies below the onset of the inelastic channel @xmath82 . in this region
the system equations become , @xmath83 when @xmath84 .
it needs to be emphasized that @xmath85 here is not an eigenvalue to be determined but it is fixed given the energy @xmath86 of the incident beam .
the open channel solutions @xmath87 correspond to continuum states while the closed or evanescent channel function @xmath88 is exponentially damped .
however , since we are concerned with the mixing between @xmath87 & @xmath88 pertaining to a bound state in the continuum we must adopt a scheme for the normalisation of the two functions treated at par .
this is indeed what is proposed by bayman & mehoke @xcite where the open channels are normalized to unit flux and the closed channels with an analogous factor ( which would result from analytic continuation ) .
thus with @xmath89 & @xmath90 being the probability amplitudes ( or mixing coefficients ) to be determined , @xmath91 which when substituted in the system equations ( [ eq : sch1nn ] & [ eq : sch2nn ] ) yield @xmath92 it may be observed that @xmath93 in view of unitarity since the inelastic channel is closed ( or evanescent ) .
note that in the limit @xmath94 we have @xmath95 which means that the bound state decouples and probability to be in the elastic channel is unity .
when we try to access the bound state embedded in the continuum through elastic scattering ( or other processes ) the transition probability will involve the overlap integral .
the probability for exciting the ` auto - ionizing ' state described via scattering is proportional to the square modulus of the overlap integral between @xmath96 & @xmath97 which is @xmath98 ^ 2\ ] ] which is seen to vanish when @xmath99 that is when @xmath100 .
this is the celebrated fano zero with the experimental signature that the line shape for excitation of the auto - ionizing state vanishes at some energy @xmath101 .
therefore , it may also be observed that the transmission amplitude , @xmath102 ( see appendix a ) has a zero at @xmath103 . to form a picture of the composite situation that has emerged it is worthwhile to look at this again for small @xmath104 whence we find that the fano - zero is located at @xmath105 $ ] as compared to the resonance energy @xmath106 $ ] and the width @xmath107 with the fano parameter @xmath108 .
it is evident from the above discussion that the essence of the fano zero lies in the mixing between the continuum states of the elastic channel and quasi - bound ` discrete ' state of the inelastic channel .
this is well expressed by ugo fano @xcite : repulsion ( of levels ) is familiar in discrete spectra , when a level of one configuration happens to lie in the midst of a series of levels of other configurations ; configuration interaction causes the levels of the second series to be shifted away from the perturbing level .
in our situation , where a discrete level of one series lies in the continuum of another channel , the levels of this continuum are also repelled , but in the sense that their oscillator strength ( level densities ) is thinned out in the proximity of the perturbing level " ( the addition under quotes given in parentheses are ours and are provided in the interest of increased clarity ) . to concretise the remark in the present context we observe the effect of the coupling on the density of states ( dos ) of the elastic channel . the changed dos due to the auto - ionizing state " is thus given by the multiplicative factor @xmath109 thus reducing the density of levels via level repulsion " .
to provide a three dimensional version of the problem at hand we consider a spherical shell type coupling potential @xmath110 $ ] between the two channels and the same form @xmath111 $ ] for the potential operative in the inelastic channel .
we confine ourselves to low energy scattering so only the @xmath112 partial wave need be taken into account . defining @xmath36 , @xmath37 , @xmath38 and @xmath39 the system equations [ vide eq .
( [ eq : sch1 ] ) & eq .
( [ eq : sch2 ] ) ] become @xmath113 + { \tilde { { \lambda}}_c } { { \delta}}(r - a ) \phi(r ) = -k^2 \psi(r)\label{sec2sch1}\\ \frac{1}{r^2 } \ , \frac{d}{dr}[r^2 \frac{d\phi(r)}{dr } ] + { \tilde { { \lambda } } } { { \delta}}(r - a)\phi(r ) + { \tilde { { \lambda}}_c } { { \delta}}(r - a)\psi(r ) = -q^2 \phi(r ) \label{sec2sch2}\end{aligned}\ ] ] using @xmath114 and @xmath115 in eq .
( [ sec2sch1 ] ) & eq .
( [ sec2sch2 ] ) , the system equations are transformed to @xmath116 we observed that when two channels are uncoupled ( i.e. @xmath47 ) , similar to previous case , the elastic channel corresponds to free particle while inelastic channel can be described by continuum solutions and a bound state . in presence of coupling between two channels we must solve eq .
( [ sec2elas ] ) & eq .
( [ sec2inelas ] ) subject to behaviour @xmath117 @xmath118 describing particle incident in the elastic channel .
note that @xmath119 and @xmath120 as required for well behaved wave functions .
using proper boundary conditions for this system ( continuity of the wave functions and jump condition for the derivatives ) we obtain @xmath121 we find that @xmath122 which is a consequence of unitarity with respect to the two channels elastic and inelastic .
as was the case for the 1d model we can see that the poles of @xmath123 has shifted in presence of the coupling between two channels . when the inelastic channel is closed ( evanescent ) i.e. @xmath82 or @xmath124 is purely imaginary ( say @xmath125 ) then @xmath126 and the bound - state of the inelastic channel can be degenerate with the continuum of the elastic channel . from eq .
( [ soln - elas ] ) and eq .
( [ soln - inelas ] ) , when two channels are uncoupled ( i.e. @xmath127 ) , we observe that @xmath128 and @xmath129 which imply that everything is transmitted out through the elastic channel .
the square modulus of the overlap integral between the closed inelastic channel @xmath96 and open elastic channel @xmath130 is @xmath131 ^ 2 sin^2(ka)\ ] ] which is seen to vanish when @xmath132 + \frac{{\tilde { { \lambda}}_c}^2 } { 2 } \left [ 1 - e^{-2 \beta a } \right]^2 = 0\ ] ] since this is a transcendental equation one should take recourse to numerical calculations , but we choose for simplicity to consider the situation where @xmath60 is small and @xmath133 is small ; hence @xmath134 is close to unity .
in such a circumstance we find that the overlap integral vanishes when @xmath135 or at an energy @xmath136\ ] ] which is the fano zero " .
to provide another soluble example , this time again in three dimensions , to illustrate the notion of the fano resonance , we introduce the idea of a non - local potential . to make this article self - contained for begining students we reconsider the schrdinger equation for a stationary state @xmath137 for a particle of mass @xmath138 moving in a potential @xmath139
is in the coordinate representation @xmath140 .
now @xmath141 where we have simply used the completeness of the states @xmath142 .
usually we deal with local potentials where @xmath143 and the potential term simply reduces to @xmath144 . however , in general , since a particle in quantum mechanics is not necessarily localized we may admit non - local potentials . furthermore ,
if we take a non - local but separable and central potential @xmath145 where we have also chosen a specific ( yukawian ) form of the radial dependence for illustration , we have several simplifications .
firstly the potential is effective only in the @xmath146 state because only that component survives the onslaught of the angular integration in @xmath147 .
secondly the solutions of the equation involve nothing more complicated than exponential functions . thus choosing a model analogous to that described by the system equations ( [ sec2elas ] ) & ( [ sec2inelas ] ) except that now we have in place of delta functions the non - local separable potentials of the chosen form
, we are invited to consider the equations @xmath148 where we have written @xmath149 and @xmath150 .
the other symbols have meanings completely analogous to what we had earlier viz .
@xmath151 , @xmath152 , @xmath153 , @xmath154 .
we have written the system equations for energies @xmath155 viz . above the threshold where both channels are open .
first putting @xmath156 we observe that the elastic channel corresponds to a free particle ( so chosen as to avoid unnecessary complications we have , as earlier , taken the elastic channel to be non - trivial only through its coupling to the inelastic channel ) .
on the other hand the inelastic channel has continuum solutions described by @xmath157 corresponding asymptotically to incoming and outgoing spherical waves ( the latter suffering a phase - shift @xmath158 ) and the last term inserted to guarantee that @xmath159 as @xmath160 $ ] and this is needed to guarantee the well - behavior of the wave function . substituting eq .
( [ eq : nonloc2sol ] ) in eq .
( [ eq : nonloc2 ] ) setting @xmath161 we get @xmath162 provided @xmath163 & the potential is sufficiently strong to support a bound state .
now let us consider the effect of the coupling between the two channels . with particle incident in the elastic channel we must solve equations ( [ eq : nonloc1 ] ) & ( [ eq : nonloc2 ] ) subject to the behaviour @xmath164 with only outgoing asymptotic waves in the inelastic channel , and of course @xmath165 and @xmath166 .
substitutions into the equations ( [ eq : nonloc1 ] ) & ( [ eq : nonloc2 ] ) yield @xmath167 again we can verify when @xmath168 is real that @xmath169 and when @xmath170 , @xmath85 real @xmath171 which are consequences of inelastic and elastic unitarity respectively . however when @xmath172 then @xmath173 . as was the case for the dirac delta model we can see that the poles of @xmath174 has shifted as compared to that of @xmath175 due to the coupling between the channels embodied by @xmath176 and one similarly finds that the bound state of the inelastic channel becomes a resonance in the elastic channel corresponding to a pole located in the fourth quadrant of the second riemann sheet in the complex energy plane .
+ in a similar fashion examining the situation at energies below the onset of the inelastic channel ( @xmath177 ) we may study the mixing of the quasi - bound auto - ionizing " state and the continuum and the probability for exciting this state via scattering to discover the energy at which the overlap integral vanishes ( _ viz . _ the fano zero ) @xmath178\end{aligned}\ ] ]
we have attempted to explain the origin of fano resonance through three soluble models : with one - dimensional dirac delta potential , with dirac - delta shell potentials and with a non - local separable potential in three dimensions . to make the results more transparent
we have taken the elastic channel to correspond to a free particle but for its coupling to the inelastic channel so that the bound state of the latter manifests itself through the typical fano pole - zero structure .
of course one could have included a direct potential in the elastic channel as well but this would only complicate the algebra without adding anything to the understanding of the relevant mechanism .
it has also been clarified how the amplitudes generate a pole in the fourth quadrant of the second riemann sheet of the complex energy plane in the elastic unitary cut , and how the excitation functions for the auto - ionizing " quasi - bound state from the elastic channel develop a zero at some real energy near the resonance .
the generic reason for the occurrence of this zero is explained by fano as arising from the fact that the wave function in the elastic channel will be of the form @xmath179 which is @xmath180 modulo an over - all phase and irrelevant factors . to find the excitation probability of the
quasi - bound state we have to determine the overlap of this with some smooth function @xmath181 [ which varies slowly with respect to @xmath61 ] .
now @xmath183 .
as the energy crosses the fano resonance @xmath184 changes from @xmath185 to @xmath186 , the overlap integral must pass through zero . in order to make the one - dimensional model to also conform to this generic pattern we have adopted the formalism of eigen - channels to introduce therein the notion of a phase - shift .
we have also demonstrated the modification in the density of states due to the bound state in the continuum " .
it is perhaps necessary to add a word of caution .
it may sometimes be very difficult from the data to discriminate between a fano - resonance & the interference effects between a breit - wigner resonance & a slowly varying background . thus consider the latter situation described by an amplitude @xmath187 and note that @xmath188 .
the interference term between the resonance & the background could very well be parametrized by the fano form with the asymmetry parameter @xmath189 .
we are grateful to dr .
prosenjit singha deo who introduced us to the idea of the fano resonance and encouraged us to search for simple illustrations to clarify the underlying mechanism .
for scattering in three dimensions from spherically symmetric potentials the @xmath190-matrix is diagonal in the orbital angular momentum partial wave channels & hence unitarity forces each diagonal to be of the form @xmath191 , where @xmath192 is the corresponding phase shift . for a one dimensional problem with particles incident from the left @xmath193 we search for solutions @xmath194 where @xmath195 is
the reflection amplitude and @xmath196 is the transmission amplitude for the left incident particles .
similarly for particles incident from the right @xmath197 we look for @xmath198 where @xmath199 is the reflection amplitude and @xmath200 is the transmission amplitude for the right incident particles .
thus the @xmath190-matrix has the form @xmath201 now if @xmath202 enjoys the symmetry @xmath203 then clearly @xmath204 and @xmath205 & accordingly as @xmath206 it possesses the eigenvectors @xmath207 and @xmath208 belonging to eigenvalues @xmath209 & @xmath210 respectively & hence with these eigenchannels as basis the @xmath190-matrix is diagonal with @xmath211 .
noting that for @xmath212 we have non - trivial scattering only in the even channel .
thus @xmath213 and @xmath214 ( in the notation of the paper ) .
therefore , @xmath215 .
it may be noted that some authors @xcite adopt a different definition of the @xmath174-matrix from that we have taken .
it is best to look at @xmath216 above @xmath217 viz .
both channels open & track down its pole in @xmath218 so that the sheet structure also becomes clear .
( [ eq : wavampl ] ) gives for the pole @xmath219 at @xmath220 the pole is at @xmath221 or @xmath222 with @xmath223 viz .
now switch on @xmath72 & assuming @xmath72 is very small , we may write @xmath225 where @xmath226 and @xmath227 are small & go smaller with smaller @xmath72 . + thus ,
@xmath228 and @xmath229 \label{eq : polea - qtru}\end{aligned}\ ] ] here in taking the square - root of @xmath230 we have carefully taken the correct sign as with @xmath231 the pole must be at @xmath232 .
@xmath233 keeping terms up to the 2nd order in @xmath234 in expressions ( [ eq : polea - qtru ] ) & ( [ eq : polea - ktru ] ) and then using them in eq .
( [ eq : polea ] ) we note that lowest terms in @xmath234 ( with @xmath81 of the same order ) in imaginary & real part of the eq .
( [ eq : polea ] ) yield @xmath235 note that while @xmath226 is of @xmath236 , @xmath227 is of @xmath237 . from above two relations @xmath238 as @xmath226 is of @xmath236 the imaginary part of the eq . ( [ eq : polea ] ) must be considered in a way so that it contains @xmath226 and @xmath239 which are of @xmath240 .
thus we get @xmath241 using expressions ( [ eq : reshift ] ) & ( [ eq : imshift ] ) we find the pole position from eq .
( [ eq : pole ] ) as @xmath242 -i \ , \frac { \hbar^2 } { 2 m } \,{\tilde { { \lambda}}_c}^2 \ , \frac{{\tilde { { \lambda}}}}{4({\tilde { { \delta } } } - \frac{{\tilde { { \lambda}}}^2}{4})^{1/2 } } \label{eq : pole - e}\end{aligned}\ ] ] which are referred as eq .
( [ eq : pole - del - k ] ) & eq .
( [ eq : pole - del - e ] ) in the text .
o. k. rice , j. chem .
* 1 * , 375 ( 1933 ) u. fano , nuovo comento * 12 * , 156 ( 1935 ) u. fano , phys . rev . * 124 * , 1866 - 1878 ( 1961 ) c.fhner et
. al . , phase measurements using two - channel fano interference in a semiconductor quantum dot , cond - mat/0307590 ( 2003 ) k.kobayashi et.al . , phys .
* 88 * , 256806 ( 2002 ) ; k.kobayashi et.al . ,
mesoscopic fano effect in a quantum dot embedded in an aharonov - bohm ring , cond - mat/0309570 ( 2003 ) b. f. bayman and c. j. mehoke , am .
* 51 * , 875 - 883 ( 1983 ) u. fano and j. w. cooper , reviews of modern physics , * 40 * no.3 , 441 - 507 ( 1968 ) s. datta , _ electronic transport in mesoscopic systems_. cambridge university press , cambridge , uk , pages 119 - 140 ( 1997 ) . | the fano resonance , involving the mixing between a quasi - bound ` discrete ' state of an inelastic channel lying in the continuum of scattering states belonging to the elastic channel , has several subtle features .
the underlying ideas have recently attracted attention in connection with interference effects in quantum wires and mesoscopic transport phenomena .
simple toy models are provided in the present study to illustrate the basics of the fano resonance in a simple and tractable setting . |
pulsed radio - frequency ( rf ) discharges are becoming an increasingly important subset of rf plasmas , particularly for industrial applications where ion energy distribution control is important .
pulsed plasmas are often used to mitigate charge build up on surface substrates .
electrons are the species responsible for this mitigation , so time resolved electron density measurements are particularly important . obtaining these time resolved measurements of transients in rf pulsed discharges can be done through a number of diagnostic methods .
however , highly time resolved measurements on the order of 100 s of nanoseconds can be considerably more difficult@xmath0}$ ] .
additionally , most diagnostic methods are accompanied by problematic drawbacks . for instance , langmuir probes perturb the local plasma by drawing current and affect measurements and have trouble working in harsh chemical environments because thin films deposited on their surface can inhibit charge collection .
interferometry suffers from a lack of spatial resolution , can often be costly , and not possible in certain experimental setups , but suffers few other drawbacks .
hairpin probes operated in boxcar mode offer nearly the same time resolution as interferometry but with the added bonus of spatial resolution and have none of the drawbacks inherent to langmuir probes .
design of a hairpin probe depends on the type of plasma source being measured .
the first choice one faces is deciding on the type of probe to employ .
there are three choices available : grounded reflection , floating reflection , and grounded transmission .
floating transmission probes are not possible since the hairpin needs to be connected to an external ( and thus grounded ) circuit .
transmission probes are a viable option and have the added benefit of not requiring a directional coupler since reflected signal is no longer being measured , but involve more complicated construction .
transmission probes tend ot have a higher q value@xmath1}$ ] , which is important for higher pressure plasmas where collisions will negatively influence the signal quality . after deciding this
, one needs estimates for the pressure ranges , electron densities , and electron temperatures .
the peak electron densities combined with the desired hairpin dimensions will yield the required maximum signal generator frequency required , as seen in equation 1 .
the hairpin dimensions need to be such that the radii of the hairpin wires must be much less than the electron mean free path and also long enough to keep the resonance frequency low .
further hairpin probe design considerations can be found in literature@xmath2}$ ] the floating probe has the added benefit of minimizing the effects of the rf sheath oscillations since it is floating with the rf potential .
the measurements taken for this work are done using a grounded reflection hairpin probe , which is the simplest to construct .
@xmath3 neglecting sheaths is valid for floating probes but a sheath correction remains important for a grounded probe in order to obtain accurate densities , as seen in equation 2 .
this correction also has a small impact on electron density rise and decay times .
the first 5 - 10 @xmath4 will particularly affected by this , since sheath thickness is dependent upon electron temperature , which changes much faster than density .
this relationship is investigated further in section iii .
since hairpin probes resonate at a characteristic frequency that depends on the permittivity of the medium and the length of the hairpin , and the frequency being used needs to be above the plasma frequency to avoid additional circuit loading , frequencies on the order of 2 - 5 ghz are typically used .
these frequencies can be difficult to work with and attenuate quickly over most cables .
this issue can be minimized by using the experimental setup given in figure 1 .
cheap wideband synthesizers are readily obtainable and can be used in conjunction with a directional coupler and a mixer to produce two heterodynes of larger and lower frequencies , given by the sum and difference of the two mixing frequencies .
figure 2 shows how the input waveform changes after each step in the circuit .
the lower frequency can be adjusted to be low enough to not attenuate and high enough to provide sufficient time resolution as needed .
the importance of having a well made probe lies in the quality factor of the circuit .
the resonance due to the hairpin can become indistinguishable from chamber or cable resonances if the quality factor is not high enough , since the circuit will be loaded when plasma is present . hairpin probes in constant waveform ( cw ) mode require a frequency sweep over the range of frequencies near the expected resonant frequency .
if this range is not known , then a broad sweep must be taken so the resonance is not missed .
these cw mode frequency sweeps must be calibrated to the vacuum case to avoid cable resonances .
it is often easier to use network analyzers in this case instead of a homemade circuit , since network analyzers readily scan multiple frequencies . for pulsed operation
, boxcar mode simplifies things greatly . in this mode , frequency is kept constant during a pulse cycle and at some point during the pulse , the density will correspond to a resonant frequency that coincides with the applied frequency .
this corresponds to a dip in the reflected signal , seen in figure 2 .
all measurements were taken at the center of a 4 @xmath5 diameter , 6 @xmath5 long cylindrical quartz housing inductively coupled source over a range of ar / o@xmath6 mixtures .
the plasma is produced inside the chamber using a 3-turn copper coil wrapped around the quartz housing , with @xmath7 cm spacing between turns , driven by a pulsing generator at 13.56 mhz .
a pi type matching network with a series variable capacitor connected with rg-8 cables is used for matching .
the details of this match are available by request .
the vacuum system included one rough pump and one turbo - molecular pump .
a pulsing signal generator is connected to the power supply and the power supply is driven in `` slave '' mode .
this allows for easy synchronization with the on - off pulse cycles .
mass flow controllers and a butterfly gate valve are also used to easily control pressure and gas content .
the parameters varied include pulse frequency , power , gas composition , and pressure .
full pulse profiles for discharges with pure ar and pure o@xmath6 are given in figures 3 and 6 , respectively .
the pulse for pure ar exhibits a typical pulsed rf profile with a time constant similar to those found in literature@xmath8[4]}$ ] .
variation of the pressure revealed a change in slope , or a `` kink '' , when pressure was plotted against decay constant , shown in figure 4 .
this kink was found to be near the same pressure the vacuum vessel transitions from the constant diffusion regime to the variable mobility regime .
the slopes for both regions are consistent with the slopes expected from both regimes .
the power was varied at a pressure of 30 mtorr , with a pulse frequency of 500 hz , duty cycle of 50% , and a flow rate of 10 sccm , as seen in figure 7 .
the decay constant is relatively independent of power , except for the region around 15 - 20 @xmath9 .
this power coincides with the e - h transition for this chamber at these settings .
a plot of the e - h transition at the same settings is shown in figure 8 .
the electron temperature at the center of an icp plasma increases dramatically when it transitions to h mode .
higher electron temperatures suggest a larger diffusion coefficient@xmath8}$ ] and thus shorter decay times , so one would expect smaller decay time constants .
however , since the electron temperature drops very quickly in the off - phase relative to the electron density , it is important to obtain an accurate estimate or measurement of the temperature .
some sources have electron temperature decay constants on the order of 10 - 16 @xmath10s@xmath11}$ ] , which would be a larger enough timescale to have a noticeable effect on the diffusion in the period that electron density decay time constants are typically taken from ( @xmath1240 @xmath4)@xmath13}$ ] .
the pulse frequency was also varied at a pressure of 30 mtorr , duty cycle of 50% , and a flow rate of 10 sccm , shown in figure 6 .
it appears that there is a smaller dependence of decay constant on pulse frequency .
the physical reason for this phenomenon is still under investigation .
the steady state on - cycle electron density changed in a nonlinear fashion as pulse frequency increased , initially decreasing and then increasing from @xmath14-@xmath15 khz .
one would expect the peak density to decrease because the time averaged power is dropping as pulse frequency is increased . th small increase from @xmath14-@xmath15 khz may be a result of the difference between the rise and decay constants . as pulse frequency increases ,
the on and off periods become shorter . since the plasma reaches a steady state value for the on - phase far quicker than the off - phase reaches steady state ,
the shorter periods keep the plasma from decaying as much .
this produces a larger steady state on - phase electron density for the proceeding cycle , resulting in larger on - phase densities after a few cycles .
this phenomenon was also noted by lieberman et al@xmath16}$ ] .
a time - averaged power measurement will have to be taken to decouple these two competing phenomena .
argon and oxygen discharges are measured over a number of different parameters including pressure , power , pulse frequency , flow , duty cycle , and ar / o@xmath6 ratios .
this work is supported by the nsf doe partnership on plasma science , the nsf goali program , and mks instruments . a special thanks to lynda larson at treasure isle jewelers for doing such an excellent job custom making our hairpin probe . | time resolved electron density measurements in pulsed rf discharges are shown using a hairpin resonance probe using low cost electronics , on par with normal langmuir probe boxcar mode operation .
time resolution of less than one microsecond has been demonstrated .
a signal generator produces the applied microwave frequency ; the reflected waveform is passed through a directional coupler and filtered to remove the rf component .
the signal is heterodyned with a frequency mixer and read by an oscilloscope . at certain points during the pulse , the plasma density is such that the applied frequency is the same as the resonance frequency of the probe / plasma system , creating a dip in the reflected signal .
the applied microwave frequency is shifted in small increments in a frequency boxcar routine to determine the density as a function of time .
the system uses a grounded probe to produce low cost , high fidelity , and highly reproducible electron density measurements that can work in harsh chemical environments .
measurements are made in an inductively coupled system , driven by a single frequency pulsing generator driven at 13.56 mhz and are compared to results from literature . |
the challenges of the lhc are to discover of the higgs boson and to search for new physics beyond the standard model ( sm ) .
the discovery of the higgs boson was reported by the atlas @xcite and the cms collaborations @xcite .
meanwhile no signals of new physics have been observed so far from the lhc .
however the nature indicates the existence of physics beyond the sm that accounts for the shortcomings in the sm , e.g. , no candidate of dark matter , the origin of neutrino mass , the baryon asymmetry of the universe , and so on . it is a challenge to confirm the new physics from only experimental / observational data .
theoretical studies in advance are essential to identify the signatures of new physics .
we now have measurement data of the higgs boson , the relic abundance of the dark matter , and so on .
what we should do , therefore , is to precisely extract probable parameter space , and predict the signatures in each scenario of new physics by using the measurement data .
one of the leading candidates of the new physics is supersymmetric ( susy ) extension .
a scenario in the extension is the constrained minimal supersymmetric sm ( cmssm ) which is simple but phenomenologically successful framework @xcite . in the cmssm ,
all of the observables are described by only five parameters , and these parameters are tightly connected with the property of the higgs boson .
the reported mass of the higgs boson , @xmath1gev , suggests heavy susy particles @xcite .
this suggestion is consistent with a null signal of exotics at the lhc .
the heavy susy particles imply the heavy dark matter in this framework , because the lightest susy particle ( lsp ) works as the dark matter .
cosmological and astrophysical measurements confirmed the existence of dark matter , and numerical simulations suggest that weakly interacting massive particles ( wimps ) are the most feasible candidate for the dark matter @xcite . in the cmssm with r - parity conservation , the bino - like neutralino is the lsp and consequently is a wimp dark matter candidate .
the measured abundance of the neutralino dark matter can be acquired in the coannihilation region .
the heavy neutralino dark matter requires large coannihilation rate .
the large coannihilation rate sufficiently reduces the relic number density of the neutralino , and can reproduce the measured abundance of dark matter . the large coannihilation rate needs the tight degeneracy in mass between the neutralino lsp and the stau nlsp ( nlsp : next to the lightest susy particle ) @xcite . indeed in large part of parameter space wherein
the mass of the higgs boson is consistent with the reported one , the mass difference between the neutralino lsp and the stau nlsp is smaller than the mass of tau lepton , @xmath2 @xcite .
such a tight degeneracy makes the stau nlsp to be long - lived charged massive particle ( champ ) @xcite .
the long - lived champs can modify the chain of nuclear reactions in a stage of big - bang nucleosynthesis ( bbn ) , and hence distort the primordial abundances of light elements .
the success and failure of the nucleosynthesis are quite sensitive to the property of the long - lived champs , e.g. , the lifetime , the number density , the electric charge , and so on @xcite .
the success of the nucleosynthesis means to reproduce the measured abundances of light elements .
notice that measured abundance of the lithium-7 ( @xmath3li ) is reported to be inconsistent with the theoretical prediction in the standard bbn ; the measured one is @xmath4
@xcite and the theoretical prediction is @xmath5 @xcite . this inconsistency is known as the lithium-7 problem @xcite , and the success of the nucleosynthesis includes also solving the problem .
it is important to emphasize that the success of the nucleosynthesis constrains and predicts the property of the stau nlsp .
the purpose of this work is to give theoretical clues to experimentally identify the cmssm as the new physics in the light of the natures of the higgs boson , the measured abundance of dark matter , and the success of the nucleosynthesis .
we concentrate on the scenario wherein the mass difference of the neutralino lsp and the stau nlsp is smaller than the mass of tau lepton , and the longevity of the stau is guaranteed by the tight phase space .
both the mass and the signal strength of the higgs boson are correlated with parameters of the stop sector .
so it sheds light on the sign(@xmath6 ) , @xmath7 , @xmath8 , and @xmath9 .
the predictability of the derived relations , however , is not so strong because of too large degrees of freedom of their parameter space . besides the property of the long - lived stau
is predicted / constrained from the nucleosynthesis , and is projected mainly on the values of @xmath7 , @xmath8 , and @xmath9 .
it should be noted that , combining the relations from the natures of the higgs boson and the values of sfermion parameters from the bbn , as we will find , the linear relation between @xmath7 and @xmath8 are derived .
furthermore the degeneracy of the neutralino lsp and the stau nlsp in the coannihilation region set a relation between @xmath10 and sfermion parameters . thus accumulating all of relations and constraints , we give theoretical clues on the parameter space of the cmssm .
then we calculate the observables of terrestrial experiments based on the analysis . after the discovery of susy signals , by checking those with the calculation we make it possible to confirm the cmssm in the near future .
this paper is organized as follows . in next section
we recall the framework of the cmssm , keeping an eye on the coannihilation scenario .
then we set the constraint on some input parameters from the viewpoint of the report of the higgs boson and the observed abundances of both dark matter and light elements . in sec .
[ sec : prediction ] , we show the predictions for susy mass spectrum , the anomalous magnetic moment of muon , branching fractions of the @xmath0-meson rare decays , and direct detection of the neutralino dark matter in the allowed region .
we show the number of the signals of the long - lived stau and the neutralino at the lhc experiment , and then we discuss the verification of the scenario in sec [ sec : lhcexp ] .
[ sec : sum ] is devoted to a summary and a discussion .
here we show our constraints to derive the experimentally favored parameter space in the cmssm . before we explain the detail of our constraints , we briefly review the cmssm .
the cmssm is described by four parameters and a sign , @xmath11 where the first three parameters are the universal scalar mass , the universal gaugino mass , and the universal trilinear coupling at the scale of grand unification , respectively . here
@xmath12 is the ratio of vacuum expectation values of two higgs bosons , and @xmath6 is the supersymmetric higgsino mass parameter . in the cmssm , we describe all observables by these parameters , and hence we can derive the favored parameter space by putting the experimental / observational constraints on the calculated observables . from now on , we explain in more details our constraints and how to apply those to our numerical calculations .
the first requirement comes from the higgs boson mass .
the latest reports on its mass , @xmath13 , are @xmath14 by the cms collaboration @xcite , and @xmath15 by the atlas collaboration @xcite , respectively .
it is known that the higgs boson mass calculated by each public codes fluctuates by about @xmath16gev @xcite .
taking into account these uncertainties , we apply a more conservative constraint as @xmath17 the second constraint comes from the observation for the relic abundance of the dark matter .
the wmap satellite reported the value at the 3 sigma level @xcite , @xmath18 in most of the cmssm parameter space , the relic abundance of the neutralino lsp is over - abundant against the measured value .
the correct dark matter abundance requires the unique parameter space where the bino - like neutralino lsp and the stau nlsp are degenerate in mass so that the coannihilation mechanism works well @xcite .
the third and fourth constraints are required from solving the lithium-7 problem .
the stau nlsp is long - lived if the mass difference is smaller than the mass of tau lepton @xcite . in the case
that the stau survives until the bbn epoch , the lithium-7 density can be reduced to the measured value with the exotic nuclear reactions induced by the stau . as the third condition
, we impose the mass difference , @xmath19 , to be @xmath20 where @xmath21 and @xmath22 are the masses of the stau nlsp and the neutralino lsp , respectively . with this mass difference , the stau nlsp is sufficiently long - lived so that it can survive until the bbn era @xcite . in our numerical calculation
, we use the pole mass of top quark as an input which involves an uncertainty of @xmath23gev .
it is expected that the susy spectrum also includes the same order of uncertainties .
so in the numerical results in the next section , we show also the case of @xmath24 to make more conservative prediction on the cmssm parameters .
the forth constraint is the upper bound on the mass of the neutralino lsp , @xmath25 this upper bound is derived from the requirement of the abundances of both the dark matter and light elements .
the yield value of the negatively charged stau at the bbn epoch is required to be larger than @xmath26 , where @xmath27 ( @xmath28 is the entropy density ) , to sufficiently reduce the lithium-7 @xcite .
assuming the enough longevity of the stau , @xmath29 is fixed at when the ratio of number densities of the stau nlsp and the neutralino lsp are frozen out .
the yield value @xmath29 is expressed with the relic density of the neutralino dark matter @xmath30 @xmath31 here @xmath32 is the freeze - out temperature of the ratio , @xmath33 . the relic abundance of the dark matter is bounded ( see eq . ( [ eq : dmabundance ] ) ) , @xmath34 here @xmath35 is the today s entropy density , @xmath36 is the scale factor of the hubble expansion rate , and @xmath37 is the critical density . combining these facts , we obtain the upper bound on the mass of the neutralino dark matter , @xmath38 hence the upper bound on the neutralino mass is @xmath39gev .
in this section , we show the allowed region in the cmssm parameter space by imposing the constraints explained in sec . [
sec : parameter region ] .
we calculate the susy spectrum , the dm abundance and other observables using micromegas @xcite implementing spheno @xcite , and the lightest higgs mass using feynhiggs @xcite .
in fig . [ fig : a0-m0plane ] , we show the allowed region in @xmath41-@xmath40 plane for @xmath42 and @xmath43 from top to bottom and @xmath44 and @xmath45gev from left to right panels , respectively .
color represents @xmath46 .
notably , one can see that @xmath40 and @xmath41 have an almost linear relation for fixed @xmath46 .
the relations can be parameterized as @xmath47 these linear relations come from the tight degeneracy in mass between the stau and the neutralino .
we explain the linear relation as follows .
firstly we note that signs of @xmath6 and @xmath40 is opposite each other due to obtaining the large higgs boson mass as we explain latter . in this scenario
we choose @xmath48 and @xmath49 , respectively . for a fixed @xmath50 , in response to an increasing of @xmath7 , the mass of the lighter stau is also increased . in order to keep the mass difference to be smaller than 1gev(0.1gev ) , @xmath51 also has to be increased .
increasing of @xmath51 makes a non - diagonal element of the stau mass matrix to be large , and hence the mass eigenvalue of the lighter stau is decreased after diagonarization to be within the small mass difference . as a result , the linear behavior in the @xmath8-@xmath7 plane is found .
one can see in fig .
[ fig : a0-m0plane ] that @xmath52 in the allowed region increases as @xmath12 increases .
this is because the soft masses of the staus are more decreased for larger @xmath9 in rg running due to the tau yukawa couplings .
the terms with tau yukawa coupling in rge decrease the soft masses in the running @xcite .
these become significant when @xmath9 is large because the tau yukawa coupling is proportional to @xmath53 .
thus larger @xmath7 is required for larger @xmath9 to obtain the stau mass satisfying with eq . or .
the upper and the lower edges of the allowed region in all panels of the fig .
[ fig : a0-m0plane ] are determined by the constraints of the correct abundances both of the dark matter and the light elements .
this can be understood as follows .
since the dark matter abundance is depend on the neutralino lsp mass as shown in eq .
( [ eq : dmabundance ] ) , the dark matter abundance gives the lower bound on the neutralino mass .
on the other hand , the light elements abundance gives the upper bound as shown in eq .
( [ eq : neutralinomass ] ) .
the bounds for the lighter stau mass is nearly same as that for the neutralino mass due to the tight degeneracy .
therefore once @xmath40 is fixed , @xmath41 can vary in range satisfying the bound for the lighter stau mass .
thus , the upper and the lower edges are determined by the constraints .
on the other hand , the right side of the allowed region is determined by the lower bound on the higgs boson mass .
we can see from the the left panel of fig .
[ fig : a0-m0plane_mh_xt ] that the higgs boson mass reaches to @xmath54gev .
the mass square of the higgs boson with one - loop corrections is given by @xmath55 , \label{eq : higgsmass } \\ & ( x _ { t } = a _ { t } - \mu \text{cot } \beta , \ \
{ \tilde t } = \sqrt{m _ { \tilde t _ { 1 } } m _
{ \tilde t _ { 2 } } } ) , \nonumber \end{aligned}\ ] ] where the first line represents the tree - level mass square and the second one is the one - loop corrections @xcite .
the tree - level contribution is simply given by the @xmath56 boson mass @xmath57 while the radiative corrections are given by the masses of the top @xmath58 , the lighter / heavier stop @xmath59 , the stop mixing parameter @xmath60 , and the vacuum expectation values of the higgs bosons , @xmath61 , here @xmath62 and @xmath63 are vacuum expectation values of up - type and down - type higgs boson , respectively .
the radiative corrections are sensitive to the stop mixing parameter .
as is well known , the second term in the bracket decreases quadratically from its maximum at @xmath64 to obtain maximum value of @xmath65 , signs of @xmath6 and @xmath40 have to be opposite each other , and we choose @xmath48 and @xmath49 in this paper .
the right panel of fig .
[ fig : a0-m0plane_mh_xt ] shows @xmath66 in the allowed region for @xmath67 and @xmath68gev .
it can be seen that @xmath65 is smaller than @xmath69 at the right side edge , and hence the one - loop corrections are not so large that @xmath70 is pushed up to the lower bound .
similarly , the left side edge is also determined by the higgs boson mass bound .
the value of @xmath71 gradually becomes large as @xmath51 increases . at a large @xmath72 point ,
the value of @xmath66 is equal to @xmath73 , and therefore @xmath70 receives the maximal loop correction . in the region where @xmath74 , the loop corrections to @xmath13 are smaller and @xmath13 decreases from the maximal value .
therefore @xmath13 is smaller than @xmath54gev again at the left side edge as can be seen in the right panel of fig .
[ fig : a0-m0plane_mh_xt ] .
l to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . ,
title="fig:",width=317 ] [ fig : tb10_cut_a0m0m12_dm1 ] to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . , title="fig:",width=303 ] [ fig : tb10_cut_a0m0m12_dm01 ] + to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . ,
title="fig:",width=313 ] [ fig : tab20_cut_a0m0m12_dm1 ] to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . , title="fig:",width=299 ] [ fig : tb20_cut_a0m0m12_dm01 ] + to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . , title="fig:",width=312 ] [ fig : tb30_cut_a0m0m12_dm1 ] to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . , title="fig:",width=299 ] [ fig : tb30_cut_a0m0m12_dm01 ] l to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure . , title="fig:",width=306 ] [ fig : tb20_cut_a0m0mh_dm1 ] to @xmath75 .
a gradation of colors represents each value .
light color indicates large value , and dark color indicates small value .
we fix @xmath76 to 20 and @xmath77 in each figure .
, title="fig:",width=303 ] [ fig : tb20_cut_a0m0xt_dm1 ]
we show in fig . [
fig : m0-m12plane ] the allowed region in @xmath7-@xmath10 plane for @xmath78 and @xmath43 from top to bottom and @xmath79 and @xmath45gev from left to right panels , respectively .
color and the contours represents @xmath8 .
figure [ fig : m0-m12plane_mh_dm ] shows the higgs boson mass in the left panel and the relic abundance of the dark matter in the right panel by color on the allowed region .
we fix @xmath12 to 20 and @xmath79gev in the panels . in fig .
[ fig : m0-m12plane ] , it can be seen that @xmath51 is larger as @xmath7 is larger while it is slightly dependent on @xmath10 .
this is because the lighter stau mass is mainly determined by @xmath7 and @xmath8 as is explained in sec .
[ sec : a0m0 ] . in the figure , we obtain @xmath80gev , at @xmath81gev for @xmath82 , @xmath83gev for @xmath84 , and @xmath85gev for @xmath86 , respectively .
these points are located at the middle of @xmath7 and the upper edge of @xmath10 in the allowed region .
the value of @xmath87 is equal to @xmath88 at these points , and hence @xmath70 receives the large loop corrections . furthermore the logarithmic term in eq .
( [ eq : higgsmass ] ) becomes large as @xmath46 increases since the stops become heavy as we show later in fig .
[ fig : mstop ] .
thus , the higgs boson mass is pushed up to 126 gev
. the higgs boson mass decreases as the parameters deviate from these points . in the left panel of fig .
[ fig : m0-m12plane_mh_dm ] , it is clearly seen that the lower bound on the higgs boson mass determines the left and the right edges in the allowed region . since @xmath89 is small at the left edge , the value of @xmath65 is smaller than @xmath75
. therefore @xmath70 receives a small one - loop correction . on the contrary
, @xmath89 is large at the right edge , and @xmath65 is larger than @xmath75 .
it means that @xmath70 receives a small one - loop correction as we explained in the previous subsection .
note that the minimum value of @xmath46 in the allowed region are different in each panel .
the minimum value is determined by the lower bound on the relic abundance of the dark matter . in the right panel of fig .
[ fig : m0-m12plane_mh_dm ] , we can see that the neutralino relic abundance reaches to the lower bound at the bottom edge of the allowed region gev
. if we collect all data , these stripes are not on the allowed region . ] .
the minimum value of @xmath10 becomes large as @xmath12 increases and/or the mass difference becomes small . the @xmath12 dependence on the relic abundance is understood as follows . in the favored parameter region , the stau nlsp consists mostly of the right - handed stau .
then the dominant contribution to the total coannihilation rate of the dark matter is the stau - stau annihilation into tau leptons via higgsino exchange for large @xmath9 @xcite .
the interactions of the stau - tau lepton - higgsino is proportional to @xmath9 , and so the large @xmath9 leads the large coannihilation rate of the dark matter .
larger @xmath9 therefore makes the relic abundance to be smaller .
the dependence of the relic number density on the mass difference is understood in terms of the ratio of number density of the stau to the neutralino . at the freeze out of total number of all of susy particles ,
the ratio is propotional to @xmath90 .
since the freeze out temperature @xmath91 is almost the same for @xmath92 and @xmath73gev as long as @xmath22 is fixed , the stau number density is relatively large for @xmath93gev .
this leads that the total coannhilation rate become enlarged and hence the relic number density of the neutralino is reduced .
such reduction of the number density can be compensated by increasing the neutralino mass .
thus , the minimum value of @xmath46 should be larger for larger @xmath9 and smaller @xmath19 to meet the lower bound for the dark matter abundance .
on the other hand , the maximum value of @xmath46 is fixed from the upper bound in eq .
( [ eq : neutralinomass ] ) .
note that the upper bound is derived by the fact that the large number density of the stau is required to reduce the lithium-7 to be the measured abundance .
this requirement forbids too large mass of the neutralino .
as there is a relation between the neutralino mass and @xmath10 , i.e. , @xmath94 , the upper bound of @xmath10 is estimated by the neutralino mass , and is fixed to be @xmath73tev in the present paper .
we emphasize here that the maximum value of @xmath46 in the allowed region is unique in the parameter space . if we do not take into account the lithium-7 problem , heavier neutralino or @xmath10 is allowed , and hence the resultant range of @xmath10 is wider .
l [ fig : tb10_cut_m0m12a0_dm1 ] [ fig : tb10_cut_m0m12a0_dm01 ] + [ fig : tb20_cut_m0m12a0_dm1 ] [ fig : tb20_cut_m0m12a0_dm01 ] + [ fig : tb30_cut_m0m12a0_dm1 ] [ fig : tb30_cut_m0m12a0_dm01 ] l [ fig : tb20_cut_m0m12mh_dm1 ] [ fig : tb20_cut_m0m12omega_dm1 ]
in this section , we show our predictions on the susy spectrum , the muon anomalous magnetic moment , rare decays of @xmath0 mesons , and the dark matter direct detections .
we calculate mass spectrum of susy particles by using spheno @xcite and that of higgs boson by using feynhiggs @xcite .
we show the mass spectra of susy particles in the allowed region .
figure [ fig : msquarksmgluino ] shows the masses of the gluino , the first and the second generations squarks with respect to the lightest neutralino mass , respectively . from top to bottom , @xmath76 is varied with @xmath96 and @xmath43 , and in the left and the right panels @xmath44 and @xmath45gev .
similarly the masses of the stop and sbottom are shown in fig .
[ fig : mstop ] , those of the neutralino , the slepton and the heavier higgs in fig .
[ fig : mneutralino ] , fig .
[ fig : mslepton ] and in fig .
[ fig : mheavyhiggs ] , respectively .
note that in all of figures we have excluded the region with @xmath97gev , which is the direct bound on the long - lived champ at the lhc @xcite .
we explain behaviors of these figures in following subsections . in the cmssm
, the gluino mass parameter @xmath98 is related to the bino and the wino mass parameter @xmath99 and @xmath100 at one - loop rge as follows , @xmath101 from this relations , we can obtain the ratio @xmath102 around the tev scale .
we can see that @xmath103 is nearly 6 times larger than @xmath104 in fig .
[ fig : msquarksmgluino ] , and @xmath105 is about twice as large as that in fig .
[ fig : mneutralino ] .
this means that the second lightest neutralino is almost the neutral wino . on the other hand , @xmath106 and @xmath107
extend above @xmath73 tev because these consist of the neutral higgsinos .
their masses are given by @xmath6 parameter that is sensitive to @xmath7 for fixed @xmath22 .
the modulus @xmath108 is determined by the ewsb conditions . at tree level
the corresponding formula for the correct ewsb leads as @xmath109 .
\label{eq : mu1}\ ] ] where @xmath110 and @xmath111 are the down - type and the up - type higgs soft susy breaking masses . for @xmath112 , eq .
is approximated as follows @xmath113 the soft mass @xmath114 is sensitive to @xmath41 . the approximate solution of the one - loop rge for @xmath114 is given by @xmath115 where @xmath116 , and @xmath117 is defined in eq .
( [ eq : linearrelation ] ) .
therefore @xmath106 and @xmath107 become large with increasing @xmath7 .
meanwhile , the mass of the cp - odd higgs boson , @xmath118 , is given by , @xmath119 thus , the masses of the heavy higgs boson and the cp - odd higgs boson are determined by @xmath108 , and these are close to @xmath120 and @xmath121 as is shown in figs .
[ fig : mneutralino ] and [ fig : mheavyhiggs ] . for the first and the second generation squarks and sleptons , the effects of the corresponding yukawa couplings are negligible in rg evolutions of their soft masses .
the soft susy breaking masses are parameterized up to the one - loop order as @xcite , @xmath122 note that the slepton masses are sensitive to @xmath41 in fig .
[ fig : mslepton ] while the squark masses are insensitive to @xmath41 in fig .
[ fig : msquarksmgluino ] . in eqs .
( [ eq : qlmass ] ) and ( [ eq : qrmass ] ) the second term is dominant , and hence we approximate these equations more roughly as follows : @xmath123 meanwhile , in eqs .
( [ eq : elmass ] ) and ( [ eq : ermass ] ) the contributions to soft mass from the first term is comparable to that of the second term .
therefore , the slepton masses are sensitive to @xmath41 .
figure [ fig : mstop ] shows the masses of the stops . unlike in the case of the other sfermions , the distributions of
stop masses are in nonlinear relation .
the spread of the distributions in the lightest neutralino mass - stop mass plane is understood as follows .
the masses of the stops in mass eigenstate are given by @xmath124 where @xmath125 and @xmath126 are the soft susy breaking masses , and @xmath127 is the stop trilinear coupling . in eq .
( [ eq : stopmassa ] ) the first term is dominant , and it is a decreasing function of @xmath128 because of the top and bottom yukawa couplings .
meanwhile , the second term in eq .
( [ eq : stopmassa ] ) is an increasing function of @xmath128 .
this is because the dominant term in the square root is @xmath129 involving @xmath127 , and the coupling @xmath127 is proportional to @xmath7 due to the relation between @xmath8 and @xmath7 in eq .
( [ eq : linearrelation ] ) .
hence , for a fixed lightest neutralino mass , stop masses are not simply determined as a function of the lightest neutralino mass , and rather spread depending on the @xmath128 .
the atlas collaboration gives the bound for the stop mass from the direct stop search at 8 tev lhc run @xcite .
note that our results of the stop mass are safely above the bound .
ll & + & + & ll & + & + & ll & + & + & ll & + & + & ll & + & + & for neutralino mass .
the upper and lower figures show the results corresponding to the cases of @xmath44 gev and @xmath130 gev , respectively .
, title="fig : " ] for neutralino mass .
the upper and lower figures show the results corresponding to the cases of @xmath44 gev and @xmath130 gev , respectively .
, title="fig : " ] we show the muon anomalous magnetic moment , @xmath132 , in the allowed region .
the latest results on @xmath132 have reported that there is a @xmath133 deviation between the sm prediction and the experimental data @xcite : @xmath134 where @xmath135 . for a fixed @xmath12 ,
lighter susy particles yield larger contributions .
the masses are light as @xmath7 and @xmath10 are small .
the small @xmath7 is , however , excluded by the constraint on the higgs boson mass as shown in [ sec : m0m12 ] .
the sizable contributions to @xmath136 are obtained in the regions of small @xmath10 .
when we take into account @xmath44 and @xmath45gev , the dominant susy contributions come from the smuon - bino like neutraino loop and the muon sneutrino - charged higgsino loop .
since we took @xmath48 , both contributions are positive . in fig .
[ fig : g-2 ] , the susy contributions in the allowed region are shown for @xmath44 and @xmath45gev in the top and the bottom panel , respectively . in the panels , @xmath9 is taken to be @xmath137 , @xmath138 , and @xmath43 .
it should be noted that the muon anomalous magnetic moment is consistent with current measurements within @xmath139 levels in this scenario . and @xmath140 .
we fix @xmath12 to @xmath141 and @xmath43 , and @xmath44gev .
the experimental values are indicated with the horizontal dotted lines .
, title="fig : " ] and @xmath140 .
we fix @xmath12 to @xmath141 and @xmath43 , and @xmath44gev .
the experimental values are indicated with the horizontal dotted lines .
, title="fig : " ] we show that the branching ratios of @xmath142 and @xmath140 are consistent with the experimental results . the first evidence for the decay @xmath143 has been discovered by the lhcb collaboration @xcite , and the branching ratio has been measured as @xmath144 it is also reported by @xcite that the other important rare decay @xmath140 is strongly suppressed as @xmath145 figure [ fig : b_physics ] shows the branching ratios of the two rare decays for @xmath44gev .
the top panel is for @xmath142 and the bottom is for @xmath146 .
one can see that the branchig ratios are within the @xmath139 in our allowed region , and hence are consistent with the experiments .
.the mass fraction of light quarks in a proton @xmath147 and a neutron @xmath148 @xcite . [ cols="<,<,<,<",options="header " , ]
we have studied the scenario of the cmssm in which the so - called lithium-7 problem can be solved via internal conversion processes with long - lived staus . for the abundance of lithium-7 to be reduced to the observed one
, we imposed the following conditions , that ( i ) the mass difference of the stau nlsp and the neutralino lsp is smaller than @xmath149gev , and ( ii ) the yield value of the stau nlsp is larger than @xmath150 .
the first condition that guarantees enough long lifetime of the stau constrains the stau mass for a fixed neutralino mass , hence the scalar soft mass @xmath7 and the trilinear coupling @xmath8 , while the second one that guarantees the sufficient reduction of the lithium-7 constrains the upper bound on the neutralino mass or the gaugino soft mass @xmath10 .
we analyzed the parameter space as well as the susy spectrum of the cmssm by taking the recent results on the higgs mass , susy searches and the dark matter abundance into account . in sec .
iii , we have shown the allowed region on the @xmath8-@xmath7 and @xmath7-@xmath10 plane for @xmath130 and @xmath73gev varying @xmath151 and @xmath43 , respectively .
we found a linear relation between @xmath8 and @xmath7 in all cases as given in eq .. the relation originates from the tightly degenerate mass of the stau and the relic abundance of the neutralino .
it was also found that the allowed region on @xmath8-@xmath7 plane shifts to higher scale as @xmath12 increases .
thus the susy spectrum is heavier for large @xmath12 . on the other hand , as seen in fig.[fig : tb20_cut_m0m12mh_dm1 ] , @xmath10 is constrained between @xmath152 and @xmath153gev for @xmath154 .
the upper bound comes from the yield value of the stau while the lower one from the small mass difference and the dark matter abundance .
thus , the conditions required to solve the lithium-7 problem play an important role to determine the allowed region of the cmssm parameters .
notably , these results naturally lead to heavy susy spectrum .
the bounds on @xmath10 lead to the stau mass between @xmath155 and @xmath156gev .
such heavy staus evade the present bound from direct searches at lhc . in figs.[fig
: msquarksmgluino]-[fig : mheavyhiggs ] , we have shown the susy spectrum and the heavier higgs mass in the allowed region .
one can see that the whole spectrum is relatively heavy so that it is consistent with the null results of the susy searches at the lhc experiment . among the spectrum , one of the important predictions
is the masses of the gluino and the @xmath73st/@xmath157nd generation squarks . in fig.[fig : msquarksmgluino ] , it was shown that the the masses are clearly correlated with the neutralino mass .
this is the direct consequences from the linear relation of @xmath8 and @xmath7 .
thus in our scenario , once one of them is determined , the others can be predicted .
it is important to emphasize here that the masses of gluino and squarks are constrained between @xmath158 and @xmath159tev for @xmath160 .
these gluino and squark masses are out of reach at @xmath161 tev lhc run but really in reach of @xmath162tev lhc run .
the expected numbers of the long - lived staus ( nearly degenerate neutralino ) to @xmath163fb@xmath164 are about @xmath165 for the light neutralino mass and @xmath166 for the heavy one , respectively . with these numbers ,
the long - lived stau and the nearly degenerate neutralino will be identified easily .
thus our scenario explains why susy has not been found yet , and at the same time predicts the early discovery of susy in the coming few years .
the other important prediction is the stop mass .
the stops are also relatively heavy in our scenario .
such stop masses gives large radiative corrections to the higgs mass , and in fact the higgs mass is pushed up to @xmath167gev in sizable regions of the parameter space .
the production cross section of such a lighter stop pair is comparable to those of the gluino and the squarks , and the expected number of the stops to @xmath163fb@xmath164 are between @xmath168 and @xmath169 .
thus we can expect that the lighter stops also will be found at the @xmath162tev run of lhc .
the susy spectrum is consistent with the present results on muon anomalous magnetic moment within @xmath170 and @xmath171 and @xmath172 within @xmath173 .
the direct detection cross section of the neutralino is from @xmath174 to @xmath175@xmath176 and is in reach of xenon @xmath73ton and lux / zep @xmath138ton . in conclusion ,
our scenario is predictive and indeed testable in the coming @xmath162tev lhc run .
when susy is the solution of the lithium-7 problem and its breaking is controlled in the cmssm framework , the time to discover susy is coming soon .
the work of y.k . was financially supported by the sasakawa scientific research grant from the japan science society .
this work was supported in part by the grant - in - aid for the ministry of education , culture , sports , science , and technology , government of japan , no .
24340044 ( j.s . ) . and no .
25003345 ( m.y . ) and no .
23740190 ( t.s . ) .
o. buchmueller , r. cavanaugh , a. de roeck , m. j. dolan , j. r. ellis , h. flacher , s. heinemeyer and g. isidori _ et al .
_ , eur .
j. c * 72 * , 2020 ( 2012 ) [ arxiv:1112.3564 [ hep - ph ] ] .
m. kadastik , k. kannike , a. racioppi and m. raidal , jhep * 1205 * , 061 ( 2012 ) [ arxiv:1112.3647 [ hep - ph ] ] .
y. okada , m. yamaguchi and t. yanagida , prog .
* 85 * ( 1991 ) 1 .
j. r. ellis , g. ridolfi and f. zwirner , phys .
b * 257 * ( 1991 ) 83 .
h. e. haber and r. hempfling , phys .
* 66 * ( 1991 ) 1815 .
j. a. casas , j. r. espinosa , m. quiros and a. riotto , nucl .
b * 436 * ( 1995 ) 3 [ erratum - ibid .
b * 439 * ( 1995 ) 466 ] [ hep - ph/9407389 ] .
j. f. navarro , c. s. frenk and s. d. m. white , astrophys .
j. * 462 * ( 1996 ) 563 [ astro - ph/9508025 ] .
d. clowe , m. bradac , a. h. gonzalez , m. markevitch , s. w. randall , c. jones and d. zaritsky , astrophys . j. * 648 * ( 2006 ) l109 [ astro - ph/0608407 ] .
v. springel , j. wang , m. vogelsberger , a. ludlow , a. jenkins , a. helmi , j. f. navarro and c. s. frenk _ et al .
_ , mon . not .
astron . soc .
* 391 * ( 2008 ) 1685 [ arxiv:0809.0898 [ astro - ph ] ] .
e. aprile _ et al .
_ [ xenon100 collaboration ] , phys . rev .
lett . * 107 * ( 2011 ) 131302 [ arxiv:1104.2549 [ astro-ph.co ] ] .
o. buchmueller , r. cavanaugh , m. citron , a. de roeck , m. j. dolan , j. r. ellis , h. flacher and s. heinemeyer _ et al .
_ , eur . phys .
j. c * 72 * ( 2012 ) 2243 [ arxiv:1207.7315 [ hep - ph ] ] .
s. profumo , k. sigurdson , p. ullio and m. kamionkowski , phys .
d * 71 * ( 2005 ) 023518 [ arxiv : astro - ph/0410714 ] .
t. jittoh , j. sato , t. shimomura and m. yamanaka , phys .
d * 73 * ( 2006 ) 055009 [ arxiv : hep - ph/0512197 ] .
k. jedamzik , phys .
d * 70 * , 063524 ( 2004 ) [ arxiv : astro - ph/0402344 ] .
m. kawasaki , k. kohri and t. moroi , phys .
b * 625 * , 7 ( 2005 ) [ arxiv : astro - ph/0402490 ] .
m. kawasaki , k. kohri and t. moroi , phys .
d * 71 * , 083502 ( 2005 ) [ arxiv : astro - ph/0408426 ] .
m. pospelov , phys .
lett . * 98 * , 231301 ( 2007 ) [ arxiv : hep - ph/0605215 ] .
k. kohri and f. takayama , phys .
d * 76 * , 063507 ( 2007 ) [ arxiv : hep - ph/0605243 ] . m. kaplinghat and a. rajaraman , phys .
d * 74 * , 103004 ( 2006 ) [ arxiv : astro - ph/0606209 ] .
r. h. cyburt , j. r. ellis , b. d. fields , k. a. olive and v. c. spanos , jcap * 0611 * , 014 ( 2006 ) [ arxiv : astro - ph/0608562 ] .
k. hamaguchi , t. hatsuda , m. kamimura , y. kino and t. t. yanagida , phys .
b * 650 * ( 2007 ) 268 [ arxiv : hep - ph/0702274 ] . c. bird , k. koopmans and m. pospelov , phys .
d * 78 * , 083010 ( 2008 ) [ arxiv : hep - ph/0703096 ] .
m. kawasaki , k. kohri and t. moroi , phys .
b * 649 * , 436 ( 2007 ) [ arxiv : hep - ph/0703122 ] .
t. jittoh , k. kohri , m. koike , j. sato , t. shimomura and m. yamanaka , phys .
d * 76 * ( 2007 ) 125023 [ arxiv:0704.2914 [ hep - ph ] ] .
k. jedamzik , phys .
d * 77 * , 063524 ( 2008 ) [ arxiv:0707.2070 [ astro - ph ] ] .
d. cumberbatch , k. ichikawa , m. kawasaki , k. kohri , j. silk and g. d. starkman , phys .
d * 76 * , 123005 ( 2007 ) [ arxiv:0708.0095 [ astro - ph ] ] .
j. pradler and f. d. steffen , phys .
b * 666 * , 181 ( 2008 ) [ arxiv:0710.2213 [ hep - ph ] ] .
m. kawasaki , k. kohri , t. moroi and a. yotsuyanagi , phys .
d * 78 * , 065011 ( 2008 ) [ arxiv:0804.3745 [ hep - ph ] ] .
t. jittoh , k. kohri , m. koike , j. sato , t. shimomura and m. yamanaka , phys .
d * 78 * ( 2008 ) 055007 [ arxiv:0805.3389 [ hep - ph ] ] .
m. kusakabe , t. kajino , t. yoshida , t. shima , y. nagai and t. kii , phys .
d * 79 * , 123513 ( 2009 ) [ arxiv:0806.4040 [ astro - ph ] ] .
m. pospelov , j. pradler and f. d. steffen , jcap * 0811 * , 020 ( 2008 ) [ arxiv:0807.4287 [ hep - ph ] ] .
m. kamimura , y. kino and e. hiyama , prog .
phys . * 121 * , 1059 ( 2009 ) [ arxiv:0809.4772 [ nucl - th ] ] .
k. kohri and y. santoso , phys .
d * 79 * , 043514 ( 2009 ) [ arxiv:0811.1119 [ hep - ph ] ] .
s. bailly , k. jedamzik and g. moultaka , phys .
d * 80 * , 063509 ( 2009 ) [ arxiv:0812.0788 [ hep - ph ] ] .
s. bailly , k. y. choi , k. jedamzik and l. roszkowski , jhep * 0905 * , 103 ( 2009 ) [ arxiv:0903.3974 [ hep - ph ] ] .
r. h. cyburt , j. ellis , b. d. fields , f. luo , k. a. olive and v. c. spanos , jcap * 0910 * , 021 ( 2009 ) [ arxiv:0907.5003 [ astro-ph.co ] ] .
t. jittoh , k. kohri , m. koike , j. sato , t. shimomura and m. yamanaka , phys .
d * 82 * ( 2010 ) 115030 [ arxiv:1001.1217 [ hep - ph ] ] .
m. kusakabe , t. kajino , t. yoshida and g. j. mathews , phys .
d * 81 * , 083521 ( 2010 ) [ arxiv:1001.1410 [ astro-ph.co ] ] .
m. pospelov and j. pradler , phys .
d * 82 * , 103514 ( 2010 ) [ arxiv:1006.4172 [ hep - ph ] ] .
m. pospelov and j. pradler , ann .
nucl . part .
* 60 * ( 2010 )
539 [ arxiv:1011.1054 [ hep - ph ] ] . m. kawasaki and m. kusakabe , phys .
d * 83 * , 055011 ( 2011 ) [ arxiv:1012.0435 [ hep - ph ] ] .
t. jittoh , k. kohri , m. koike , j. sato , k. sugai , m. yamanaka and k. yazaki , phys .
d * 84 * ( 2011 ) 035008 [ arxiv:1105.1431 [ hep - ph ] ] .
m. kusakabe , a. b. balantekin , t. kajino and y. pehlivan , phys .
b * 718 * ( 2013 ) 704 [ arxiv:1202.5603 [ astro-ph.co ] ] .
k. kohri , s. ohta , j. sato , t. shimomura and m. yamanaka , phys .
d * 86 * ( 2012 ) 095024 [ arxiv:1208.5533 [ hep - ph ] ] .
m. kusakabe , k. s. kim , m. -k .
cheoun , t. kajino and y. kino , arxiv:1305.6155 [ astro-ph.co ] .
[ atlas collaboration ] , atlas - conf-2012 - 170 .
b. c. allanach , econf c * 010630 * , p319 ( 2001 ) [ hep - ph/0110227 ] .
a. djouadi , hep - ph/0211357 .
b. c. allanach , s. kraml and w. porod , jhep * 0303 * , 016 ( 2003 ) [ hep - ph/0302102 ] . b. c. allanach , a. djouadi , j. l. kneur , w. porod and p. slavich , jhep * 0409 * , 044 ( 2004 ) [ hep - ph/0406166 ] .
w. porod , comput .
commun .
* 153 * ( 2003 ) 275 [ hep - ph/0301101 ] .
w. porod and f. staub , comput .
commun .
* 183 * ( 2012 ) 2458 [ arxiv:1104.1573 [ hep - ph ] ] .
s. heinemeyer , w. hollik and g. weiglein , comput .
commun .
* 124 * ( 2000 ) 76 [ hep - ph/9812320 ] .
s. heinemeyer , w. hollik and g. weiglein , eur .
j. c * 9 * ( 1999 ) 343 [ hep - ph/9812472 ] .
g. degrassi , s. heinemeyer , w. hollik , p. slavich and g. weiglein , eur .
j. c * 28 * ( 2003 ) 133 [ hep - ph/0212020 ] .
m. frank , t. hahn , s. heinemeyer , w. hollik , h. rzehak and g. weiglein , jhep * 0702 * ( 2007 ) 047 [ hep - ph/0611326 ] .
g. w. bennett _ et al . _
[ muon g-2 collaboration ] , phys .
d * 73 * ( 2006 ) 072003 [ hep - ex/0602035 ] .
k. hagiwara , r. liao , a. d. martin , d. nomura and t. teubner , j. phys .
g * 38 * ( 2011 ) 085003 [ arxiv:1105.3149 [ hep - ph ] ] .
[ lhcb collaboration ] , phys .
* 110 * ( 2013 ) 021801 [ arxiv:1211.2674 [ hep - ex ] ] . y. amhis _ et al . _
[ heavy flavor averaging group collaboration ] , arxiv:1207.1158 [ hep - ex ] .
g. jungman , m. kamionkowski and k. griest , phys .
* 267 * ( 1996 ) 195 [ hep - ph/9506380 ] .
e. aprile _ et al .
_ [ xenon100 collaboration ] , phys .
* 109 * ( 2012 ) 181301 [ arxiv:1207.5988 [ astro-ph.co ] ] . | we explore the coannihilation region of the constrained minimal supersymmetric standard model ( cmssm ) being consistent with current experimental / observational results .
the requirements from the experimental / observational results are the 125gev higgs mass and the relic abundances of both the dark matter and light elements , especially the lithium-7 .
we put these requirements on the caluculated values , and thus we obtain allowed region .
then we give predictions to the mass spectra of the susy particles , the anomalous magnetic moment of muon , branching fractions of the @xmath0-meson rare decays , the direct detection of the neutralino dark matter , and the number of susy particles produced in 14tev run at the lhc experiment . comparing these predictions with current bounds ,
we show the feasibility of the test for this scenario in near future experiment . stupp-13 - 215 , kek - th-1667 0.15 in |
a basic building block of a superconducting quantum bit is the josephson junction , which allows coherent cooper - pair tunneling .
this josephson current provides the necessary non - linear element in an electric circuit and enables reduction of the externally controllable quantum dynamics to involve only two eigenstates @xcite .
a drawback for superconducting qubits has been that these systems can be very sensitive to various effects in their nearby environments .
the influence of two major decoherence sources , charge fluctuations and two - level fluctuators , have been substantially reduced in recent years .
qubits that are less sensitive to changes in background charge have been build @xcite , while fluctuators have been removed by decreasing the junction size .
recently , transmons and persistent - current flux ( p - flux ) qubits with extremely long decay times have been demonstrated @xcite .
it is therefore of great interest to investigate mechanisms of decoherence that were previously unobservable .
temperature dependent decay time measurements in phase@xcite and transmon qubits@xcite suggest that non - equilibrium quasiparticles might have now become the main factor limiting the decay times of superconducting qubits . in this work
we analyze qubit decay and dephasing due to quasiparticle tunneling . as the source of tunneling we consider the two following deviations from the equilibrium bcs state of superconductors .
the first one is a non - equilibrium distribution of quasiparticles , which corresponds to a finite quasiparticle density above the gap , present at all temperatures @xcite .
non - equilibrium quasiparticles could originate in a quasiparticle diffusion from higher temperature regions in an experiment , or through stray radiation , and are relatively long lived due to slow quasiparticle recombination rate @xcite .
the second process we consider is quasiparticle tunneling by breaking of a cooper pair , which can occur at typical energies of superconducting qubits if there is an extended quasiparticle density inside the gap .
it has long been observed that the number of states inside the gap is vastly larger than the predictions of bcs theory @xcite .
this effect is especially pronounced in niobium , a material that is widely used in the production of qubits @xcite . nonetheless , aluminum based superconducting devices are observed to posses sub - gap states too @xcite , but with considerably smaller density as niobium .
.comparison of decay times obtained by assuming a non - equilibrium quasiparticle density @xmath0m@xmath1 ( second column , aluminum ) or a dynes parameter @xmath2 ( third column , niobium ) .
we used qubit parameters specified in each of the citations at each qubit .
we also assume @xmath3 @xmath4ev and the validity of the ambegaokar - baratoff relation .
typical sub - gap density of states in aluminum ( @xmath5 ) does not limit the coherence here . [
cols="<,^,^",options="header " , ] our results are that the two quasiparticle processes result in a similar type of qubit decay rate .
this provides a clear comparison between the magnitudes of decoherence due to the two sources . for conventional aluminum qubits
non - equilibrium quasiparticles should have a more significant impact on decoherence , for niobium qubits the sub - gap density of states .
we also find that the sensitivity to quasiparticle tunneling is vastly different for differing qubit designs . by using the
lowest reported density for non - equilibrium quasiparticles in aluminum @xcite we obtain p - flux qubit decay times similar to recent experiments @xcite , whereas for other qubit types the results are from two to three orders of magnitude greater than the observed ones . for niobium p - flux qubits
we also reproduce usual experimental decay times by assuming a dynes - type of quasiparticle sub - gap states , observed in various other experiments @xcite .
these central results are summarized in table [ tab_comparing_decay_times ] .
the paper is organized as follows . in section [ sec_foundations ]
we introduce the hamiltonian and methods describing the qubit - quasiparticle dynamics .
the section [ sec_decoherencerates ] is devoted to discussion of decoherence processes and the corresponding rate equations .
we also discuss decay rates for single - junction qubits obtained by using typical experimental parameters of the quasiparticle environments . in section [ sec_multi_junction_flux_qubit ]
we generalize the treatment to multi - junction qubits , where a special concentration is given to flux qubits in their different designs . in section [ sec_conclusion ]
we summarize the results .
our starting point is the division of the total hamiltonian into three parts , @xmath6 here the qubit hamiltonian @xmath7 describes the collective degrees of freedom that constitute our two - level system .
the quasiparticle hamiltonian @xmath8 describes the electronic degrees of freedom in the superconductors , treated independently of the collective part .
interaction between these two parts emerges due to single - electron tunneling modeled by the tunneling hamiltonian @xmath9 .
it describes , e.g. , quasiparticle tunneling across the josephson junction which can cause transitions between the eigenstates of the qubit . in the following we discuss in more detail the central properties of each of the three parts . as the general single - junction qubit hamiltonian
we use the following sum of capacitive , inductive , and josephson coupling energy @xmath10 here @xmath11 is the number operator of electron charges on the junction capacitor @xmath12 .
it is a conjugated variable to the phase difference across the junction @xmath13 , satisfying @xmath14=i$ ] .
this choice of qubit hamiltonian corresponds to the phase and dc - flux qubits , i.e. , when the junction is placed in a superconducting loop with inductive energy @xmath15 and thread by flux @xmath16 . for the transmon one has @xmath17 and periodic commutation relations @xmath18=\pm
e^{\pm i\varphi}$ ] , related to the @xmath19-quantization of the charge that tunnels across the josephson junction .
all qubits we consider work in the limit where the josephson coupling energy @xmath20 is larger than the single - electron charging energy @xmath21 .
generalization to multi - junction qubits is straightforward , see section [ sec_multi_junction_flux_qubit ] .
the quasiparticle degrees of freedom in superconducting leads are modeled by the hamiltonian @xmath8 . for all further calculations
the quasiparticle properties are assumed to be independent of the qubit state .
we consider here two models as two separated causes of quasiparticle tunneling at energies well below the superconducting gap @xmath22 .
we label these as model @xmath23 and model @xmath24 .
signatures of the electronic structure in the two models are found from correlation functions of type , @xmath25 here @xmath26 is an electron annihilation ( creation ) operator of the state @xmath27 with spin @xmath28 ( @xmath29 ) .
the fourier transforms of the correlation functions are related to spectral and distribution functions as @xcite @xmath30 here @xmath31 and @xmath32 . in equilibrium
@xmath33 $ ] .
we define the normalized density of states @xmath34 and the pair density @xmath35 as @xmath36 here @xmath37 is the density of states nearby the fermi surface , including spin .
these functions , added with the distribution function @xmath38 , have their own characteristic forms in the two models and are all we need to know about the quasiparticle environments in the final forms of the qubit decoherence rates ( see section [ sec_decoherencerates ] ) .
in model @xmath23 we assume that in each lead quasiparticles can be described by the bcs hamiltonian but exist with a general ( non - equilibrium ) probability @xmath38 . in the bcs model
the quasiparticle and pair densities have the form @xmath39 an important quantity is the total density of quasiparticles @xmath40 in calculations @xmath41 is assumed to be a given constant .
this model has also been investigated in refs . .
as the other source of quasiparticles we consider the presence of quasiparticle states below the gap .
we assume that the densities can be expressed in the same form as for the bcs theory , but with the following broadening , @xmath42 here @xmath43 is the so - called dynes parameter @xcite and @xmath44 is a possible imaginary part of the superconducting gap @xcite .
we consider @xmath45 and @xmath46 to be given parameters and independent of energy .
the effect of these modifications of the bcs theory are very similar nearby the gap , but differ at low energies @xmath47 ( see section [ sec_lowenergyapproximation ] ) .
the broadening through an imaginary part of the superconducting gap @xcite , @xmath46 ( @xmath48 ) , is supported by the microscopic eliashberg theory .
however , a constant imaginary part of the energy , @xmath45 ( @xmath49 ) , usually fits better to experiments done with , e.g. , niobium @xcite . for all further purposes
we will only consider finite @xmath45 and @xmath49 .
the interaction between the qubit and the quasiparticle environments occurs due to the possibility of quasiparticles to tunnel across the junction(s ) and simultaneously change the charge configuration in the qubit space .
this is described by the _ modified
_ tunneling hamiltonian @xmath9 @xmath50 where the _ bare _ tunneling hamiltonian has the form @xmath51 here the states @xmath27 and @xmath52 belong to the opposite sides of the junction .
we assume a constant tunneling matrix element @xmath53 , related to the normal state tunnel resistance as @xmath54 .
the charge - transfer operator @xmath55 accounts for the corresponding changes in the charge number of the leads .
cooper - pair tunneling , which is a second - order process in quasiparticle tunneling , is already included in the qubit hamiltonian in an approximative way as the operator @xmath56 . therefore we need to subtract it from @xmath57 to avoid double counting .
the second - order expansion of the qubit s density - matrix equation of motion in the tunneling hamiltonian leads to contributions describing qubit decoherence and parameter renormalization . here
we discuss in detail the decoherence terms , i.e. , terms leading to energy decay and pure dephasing .
we then estimate their magnitude for usual single - junction qubits .
an exact derivation of the rates , a formulation for the energy - level renormalization effects , and a discussion for the effect of higher - order terms is given in the appendix .
the system we consider is a two - level system interacting with a fermionic bath .
the coupling to the bath can be divided into two parts : ( i ) the part causing transitions , @xmath58 , and ( ii ) the part causing pure dephasing , @xmath59 .
the decay rate , @xmath60 , is a result of coupling to @xmath61 .
this leads also to the ordinary dephasing rate @xmath62 .
fluctuations through @xmath63 lead to the pure dephasing rate @xmath64 , which can be interpret to be a result of low - frequency fluctuations in the qubit energy splitting due to coupling to the bath @xcite . in leading order
we obtain two distinct types of tunneling processes causing decoherence .
the first one is tunneling of an existing quasiparticle . in the case of the bcs state
this is described by an operator of the type @xmath65 this process contributes mainly in model @xmath23 .
the second one is the breaking of a cooper pair , described by the bcs operator of the type @xmath66 this process contributes only in model @xmath24 .
important here is that the operators @xmath67 are superpositions of two electron - tunneling directions , @xmath68 and @xmath69 , as long as @xmath70 .
as the qubit states are superpositions of different charge states , the direction of electron tunneling becomes indistinguishable and the two tunneling processes interfere .
this gives rise to a phase dependence in quasiparticle tunneling , similar to the @xmath71-term in the classical josephson effect @xcite .
this interference effect has also an important role in decoherence of superconducting qubits , as discussed below . for the qubits considered in this work the qubit decay rate
can be written in the form @xmath72\delta(\omega-\omega'-\delta e ) .
\end{split}\ ] ] here @xmath73 is the energy - level splitting of the qubit and the angle @xmath74 is defined as @xmath75 the matrix elements depend on the choice of superconducting qubit , but as we will discuss , they are in fact very similar for broad classes of qubits .
expressions such as ( [ eq_gamma12_defined ] ) are often derived in the bcs excitation picture , where integration over only positive ( excitation ) energies emerges . in this semiconductor - type presentation
negative energies appear as well .
but the result of the two presentations is the same .
the positive integration region of both frequencies @xmath76 and @xmath77 corresponds to tunneling of an existing quasiparticle to one direction , as the negative integration region to the other region .
the contribution with positive @xmath76 but negative @xmath77 describes breaking of a cooper pair during the single - electron tunneling .
it can exist only if there is an extended states below the gap ( because @xmath78 ) .
the contribution with positive @xmath77 but negative @xmath76 contributes only if @xmath79 and would correspond to recombination of two quasiparticles with excitation of the qubit . to simplify equation ( [ eq_gamma12_defined ] ) we assume @xmath80 and that the width of the quasiparticle distribution ( @xmath81 ) is much smaller than the qubit splitting , @xmath73 .
this means that for model @xmath23 the distribution @xmath38 is nonzero above the gap ( @xmath82 ) practically only in a very narrow region [ we have always @xmath83 .
on the other hand , for the model @xmath24 with dynes broadening ( and @xmath49 ) one can approximate @xmath84 for a finite imaginary part @xmath46 and @xmath48 one would have @xmath85 and @xmath86 .
however , in the following we assume a dynes - type broadening @xcite when considering model @xmath24 . in the discussed low - energy approximation
we obtain the following common form for the decay rates , @xmath87 important here is that the rates are proportional to @xmath88 , defined for the two cases as @xmath89 the form @xmath90 for the decay rate ( [ eq_mainresult ] ) is a result of the interference effect in quasiparticle tunneling @xcite ( see the discussion in section [ processes ] ) .
its magnitude and sign are given by @xmath91 for model @xmath23 we have @xmath92 whereas for model @xmath24 @xmath93 .
physical interpretation for this is that the subgap states in model @xmath24 are close to metallic states and do not show significant pair correlations ( @xmath94 ) , needed for the effect . in the leading - order expansion of the qubit s time evolution
the pure dephasing appears as an extra decay of coherent oscillations through transition terms proportional to @xmath95 or @xmath96 .
the pure dephasing rate is then similar to the rate ( [ eq_gamma12_defined ] ) by setting @xmath97 , and changing the matrix element , @xmath98\delta ( \omega-\omega')\ , , \\ m_2 ^ 2&= \left\vert \langle\uparrow\vert \hat t \vert\uparrow\rangle-\langle\downarrow\vert \hat t \vert\downarrow\rangle\right\vert^2 .
\end{split}\ ] ] here we have assumed that @xmath99 which is the case for most of the qubits considered in this paper .
the only exception is the flux qubit away from its symmetry point , discussed more detailed in section [ sec_multi_junction_flux_qubit ] . the value of the integration in eq .
( [ eq_mainresult2 ] ) depends on the distribution @xmath38 , which depends on the model used . in model
@xmath24 we assume the equilibrium distribution @xmath100 $ ]
. then the pure dephasing corresponds to fluctuations originating from tunneling of thermalized ( sub - gap ) quasiparticles . for this case
the integration can be easily evaluated giving @xmath101 .
this means that for small temperatures @xmath102 the pure dephasing rate is much smaller than the decay rate . in model
@xmath23 the situation is more difficult due to the logarithmic divergence of the energy integral at the energy gap .
similar singularities appear also in other properties of superconductors @xcite , but stay finite due to finite lifetime effects or gap anisotropy .
it is now crucial , which type of broadening is assumed .
if one introduces a small imaginary part @xmath46 and accordingly redefines @xmath41 in eq .
( [ eq_n_qp_defined ] ) , then the integral in eq .
( [ eq_mainresult2 ] ) remains small . as a result of this
the pure dephasing rate stays small compared to the energy decay rate .
other methods for circumventing the divergence can produce larger results for the integral .
however , as we will discuss in the next section , pure dephasing is additionally suppressed by small tunneling matrix elements .
we will now discuss the qubit part of the decoherence rates in the case of single - junction qubits .
specifically we consider the phase qubit , the transmon , the dc - flux qubit , the strongly anharmonic phase qubit @xcite and the fluxonium . strictly speaking
the fluxonium is not a single junction qubit , but we consider the integrated josephson junction array simply as an inductance . for
a detailed discussion of the fluxonium including quasiparticle tunneling in the junction array see ref . .
the discussion can be structured along two major properties .
one is the squared magnitude of the qubit matrix element in eq .
( [ eq_mainresult ] ) or of the difference between matrix elements in eq .
( [ eq_mainresult2 ] ) , divided by the tunneling resistance @xmath103 .
the second is the phase difference @xmath74 in the interference ( @xmath104 ) term . when it comes to estimating the matrix elements qubits fall broadly in two classes .
one class of qubits have eigenstates located in a single local minimum of the qubit s potential energy , @xmath105 the lowest eigenstates are then similar to that of a harmonic oscillator and they are symmetric or antisymmetric around the local minimum . here
the decay and dephasing elements are in a good approximation given by @xmath106 in this class we have clearly the transmon and the phase qubit . a qubit with a quartic potential as proposed in ref .
( strongly anharmonic phase qubit ) has the same matrix element and despite its rather unusual potential , the fluxonium belongs to this group as well for @xmath107 .
the qubit that bucks the trend is the dc - flux qubit , which in fact is very similar to the fluxonium for @xmath108 .
with its double well potential it is very different from all the other qubits .
we will discuss the form of the matrix elements in detail in the next section , but at the symmetry point they can be shown to have the approximative form @xmath109 where @xmath110 is the solution to the transcended equation @xmath111 and @xmath112 the decay matrix element is the largest of the discussed qubits .
important is also that this qubit has relatively low tunneling resistance @xmath103 .
depending on the interfence effect it can be rather sensitive to quasiparticle tunneling .
the qubits have various interference angles , as defined in eq .
( [ eq_angledefinition ] ) . for the qubits that are similar to a harmonic oscillator
this angle is given by the position of the local minimum .
the potentials of the transmon and fluxonium qubits are symmetric around the phase @xmath113 , whereas the potential of the phase qubit is also practically symmetric , for certain @xmath114 . in the dc - flux qubit
( with @xmath108 ) the interference angle is given by @xmath115 , and the situation is the same for the stronlgy anharmonic phase qubit .
this provides protection against qubit decay due to non - equilibrium quasiparticles , as then the interference between two electron tunneling directions is destructive .
however , this protection is only partial because for typical qubit splittings @xmath116 one has @xmath117 ( @xmath118 ) .
the common decay rate ( [ eq_mainresult ] ) shows that the two quasiparticle sources produce a similar type of decay rate .
it is now easy to estimate which source should be dominant in typical experimental conditions .
a crucial quantity is the dimensionless parameter @xmath119 , defined in eq .
( [ eq_maindensity ] ) .
the division by the energy gap also accounts for changes in eq .
( [ eq_mainresult ] ) due to change in the tunneling resistance , if the same qubit ( with same @xmath20 ) is build from , for example , niobium instead of aluminum , as one has @xmath120 . for an aluminum superconductor
the parameter is usually measured to have a value in the range @xmath121 ( non - equilibrium quasiparticles ) or @xmath122 ( sub - gap density of states ) .
this means that for aluminum non - equilibrium quasiparticles have usually an impact that is several orders of magnitudes larger than that of sup - gap states . on the other hand , for niobium using the typically observed dynes parameter @xcite we obtain values @xmath123 , corresponding to the upper bound values for non - equilibrium quasiparticles in aluminum . in table
[ tab_comparing_decay_times ] we compare decay times for various qubits calculated by assuming the lowest reported value for the density of non - equilibrium quasiparticles in aluminum , given in ref . . for single - junction qubits ( transmon and phase )
we find decay times of the order of ms , which is two to three orders of magnitude higher than the best experimental observations .
for the same qubits but build from niobium the decay rate , eq .
( [ eq_mainresult ] ) , simplifies to @xmath124 being consistent with the result of the classical limit . in this case
we find decay times of the order of @xmath125 @xmath4s .
the single - junction qubit that is the most sensitive to both quasiparticle sources is the dc - flux qubit , for which we obtain decay times of order 100 @xmath4s ( al ) and 100 @xmath34s ( nb ) ( not listed in table [ tab_comparing_decay_times ] ) .
the value for aluminum includes the protection due to destructive interference which is approximately a factor of @xmath125 .
all together , in these conditions the aluminum based single - junction qubits are protected against non - equilibrium quasiparticles , most of them due to small @xmath126-ratio .
in this section we generalize our analysis to multi - junction qubits .
we consider in detail the most common of them , the flux qubit , which comes in many different shapes and forms .
the original proposal , fig . [ fig_the_flux_qubit](a ) , is a qubit with three junctions @xcite , but in general it is today used in many variations often with several extra junctions . for technical reasons
the flux qubits have always been build with a fourth junction , fig .
[ fig_the_flux_qubit](b ) , which in most modern flux qubits is of the same size as the two large junctions . in another version of the flux qubit ,
the small junction has been replaced with a squid , meaning by a loop with two junctions , fig .
[ fig_the_flux_qubit](c ) .
this allows for more tunability . we will start this section with a discussion of the generalization of the decoherence rates to many junctions
. then we will continue with calculating the rates for the p - flux qubit with three junctions and then later discuss the modification by the fourth junction and the tunable third junction .
we end the section with the capacitively shunted flux qubit , which by its properties is shown to be similar to an anharmonic oscillator . as before we define our total hamiltonian of the qubit and the quasiparticle environment as @xmath127 where @xmath7 is the hamiltonian of the flux qubit , coupled via the tunneling hamiltonian @xmath9 to a quasiparticle environment @xmath8 . with a total of @xmath34 junctions we have @xmath128 , where the bare tunnel hamiltonian is @xmath129 here @xmath130 and @xmath131 are electron annihilation ( creation ) operator of the states @xmath27 and @xmath52 on the two sides of the junction @xmath132 with josephson coupling @xmath133 , and @xmath134 is the corresponding qubit - space charge - transfer operator , given in different cases below .
the quasiparticle environments in each of the leads are assumed to be identical and described by relations ( [ eq_greendefined]-[broadening2 ] ) . in the leading order
the decay rates can be written into a similar form as for the single - junction case , eq .
( [ eq_gamma12_defined ] ) , and in the low - energy approximation ( section [ sec_lowenergyapproximation ] ) one gets @xmath135 here @xmath136 and @xmath137 stand for the tunneling resistance and the interference angle of the junction @xmath132 .
similar generalization applies also for the pure dephasing rate , eq .
( [ eq_mainresult2 ] ) . in the following
we discuss the form of the matrix elements @xmath138 and the phases @xmath137 in the considered flux qubit realizations .
the first persistent - current flux qubit we consider has three josephson junctions , see fig .
[ fig_the_flux_qubit](a ) .
its hamiltonian can be presented in the form @xmath139 .
\end{split}\ ] ] here @xmath140 is the total charging energy of the system , @xmath20 is the josephson energy of the larger junctions and @xmath141 is a parameter that determines the energy splitting of the qubit . in the following it
is assumed that for the corresponding capacitive term it holds @xmath142 .
the ratio of josephson and charging energy for the p - flux qubit is generally given by @xmath143 .
we have three junctions with phase differences @xmath144 , @xmath145 and @xmath146 , as illustrated in fig . [ fig_the_flux_qubit](a ) .
the phases in the hamiltonian are then defined by @xmath147 for each dynamical phase we have a conjugate charge variable , @xmath148=i$ ] , and for each junction we have a tunneling operator given by @xmath149 the potential of the p - flux qubit is given by @xmath150 in the direction of @xmath151 the system behaves similar to a harmonic oscillator that remains in the ground state .
therefore we will focus on the dynamics in the @xmath152 direction .
choosing @xmath108 ( the symmetry point ) and @xmath153 we expand @xmath154 for small @xmath152 , @xmath155 \varphi_+^4.\ ] ] we have @xmath156 .
this means that the eigenstates of the qubit are formed by symmetric and antisymmetric superpositions of the ground state of two wells .
these ground states are centered around the minima @xmath157 , @xmath158 in the left well the eigenstate is of the form @xmath159,\ ] ] where we defined @xmath160 the state in the right well , @xmath161 , has the same form , just centered at @xmath162 . under the same approximation we used to derive the potential given by eq .
( [ eq_u_flux_qubit_approximated_in_lowest_order ] ) we can now find the coupling strength @xmath53 between the two states , which is given by @xmath163 e_j } { 1 + 2 \alpha}\nonumber\\ & & \times \exp(-\omega'\varphi_{+,{\rm min}}^2)\nonumber .
\end{aligned}\ ] ] here we assumed @xmath108 and consider the contribution of the order of @xmath20 .
there is an additional contribution of the order of @xmath164 and @xmath140 , but they remain small . often the coupling strength has been calculated using the wkb approximation .
our calculations underestimates the coupling strength but still gives a good order of magnitude approximation .
we move away from symmetry like @xmath165 and in the lowest order of @xmath166 we get @xmath167 if we now compare the energy of the minima with the energy of the minima for @xmath168 we find the energy difference @xmath169 now we can write the hamiltonian in the standard way of a two - state approximation for the states @xmath170 and @xmath161 , @xmath171 the basic form of the states is therefore given by @xmath172 with @xmath173 .
let us now consider an operator of the form @xmath174 we get in a good approximation @xmath175 and @xmath176 using these equations we will be able to calculate the effect of quasiparticle tunneling on the flux qubits . to estimate the qubit decay rate we start from the transition element of the operator @xmath177 .
this can be expressed in the form @xmath178 from this general form we can deduce the results for the tunneling operators @xmath179 and @xmath180 , @xmath181 where we have used the fact that the system stays in the ground state in the direction @xmath151 .
important here is that for the symmetry point one has @xmath182 and the matrix elements ( [ eq_pfluxelement1 ] ) are purely imaginary .
this is in contrast to the ( single - junction ) dc - flux qubit where the element is real .
it follows that the interference angles in eq .
( [ eq_mainresult3 ] ) are @xmath183 , i.e. , one has constructive interference instead of destructive .
the junction that is somewhat different is the second junction , for which the matrix element is given by @xmath184 e^{i\varphi_{\rm ext}/2}.\ ] ] the element ( [ eq_pfluxelement2 ] ) has an additional phase factor , @xmath185 , and becomes real at the symmetry point @xmath108 .
this corresponds to the case of the dc - flux qubit and destructive interference .
the interference effect has an important role in determining the sensibility to non - equilibrium quasiparticle tunneling . as discussed in section [ discussionforsinglequbits ] ,
the decay matrix element of this qubit type is the largest , but the single junction dc - flux qubit is protected by destructive interference . as a result of constructive interference for two junctions of the p - flux qubit
, this protection is lost .
it follows that even with the smallest observed non - equilibrium quasiparticle density @xcite , corresponding to @xmath186 , we obtain a decay time of the order @xmath125
this result coincides with recent experiments @xcite .
results for niobium p - flux qubits is similar to dc - flux qubits : using typical subgap densities@xcite one obtains decay times of the order @xmath187
@xmath34s , being also similar to recent experiments @xcite .
these results are compared with the ones of other qubit types in table [ tab_comparing_decay_times ] .
the dephasing is characterized by the diagonal matrix elements of @xmath134 , which can be shown to have the form @xmath188 the second junction has again an extra phase dependence , @xmath189 .
\end{split}\ ] ] the presence of a term that does not change sign when changing the state leads effectively to a different angle @xmath190 for the two states @xmath191 and @xmath192 , and deviations from result ( 19 ) .
however , near the symmetry point such corrections stay small and one can neglect this effect . in this region the main contribution to
pure dephasing comes from the junction 2 . using eq .
( [ eq_pfluxdephasing2 ] ) one obtains then for the relevant dephasing matrix element @xmath193 this is usually much smaller than the decay element .
combined with the discussion of section [ discussionforsinglequbits ] , this means that at the symmetry point the dephasing is limited only by the qubit decay . for technical reasons
the p - flux qubit is in fact always build with four junctions instead of the necessary three .
the fourth junction has a charging energy @xmath194 and a josephson energy @xmath195 .
we can assume @xmath196 .
the additional junction adds another dimension to the problem , and we can approximate states along this dimension as states of a harmonic oscillator . within this approximation
the total hamiltonian can be written as @xmath197 where @xmath198 are the pauli matrices acting on the eigenstates of the p - flux qubit ( [ eq_eigenstates_of_p_flux_qubit ] ) , with the energy splitting @xmath199 .
these states are coupled with the coupling strength @xmath200 to the fourth junction , modeled as a harmonic oscillator with frequency @xmath201 .
the coupling is determined by the following matrix elements , @xmath202 if the fourth junction is large , @xmath203 , the coupling goes to zero . in this case
the fourth junction is decoupled and always stays in the ground state .
however , in many flux qubits the fourth junction has the josephson energy @xmath204 . in this case
the coupling is strong , @xmath205 . the tunneling operator for the fourth junction is simply given by @xmath206 .
close to the symmetry point we have @xmath207 and additionally we assume @xmath208 . in this case
we can approximate the matrix element in the lowest order of @xmath209 by @xmath210 where we have labeled the two lowest eigenstates of the four - junction flux qubit as @xmath211 and @xmath212 . in the limit of weak coupling @xmath213 , we find @xmath214
. then the matrix element simplifies to @xmath215 .
in general we have @xmath216 and therefore transitions due to tunneling across the fourth junction are well suppressed compared to tunneling across the other junctions .
similarly we can calculate the tunneling through the third junction with the tunneling operator @xmath217 and , using the same approximation as for eq .
( [ eq_t4_in_maximum_approximation ] ) , we get @xmath218 in conclusion we can say that the fourth junction should have little effect on the overall decoherence rate of the p - flux qubit . in the tunable gap flux qubit the small junction is replaced by a squid , see fig . [
fig_the_flux_qubit](c ) . in this case
the parameter @xmath28 can be tuned by changing the field @xmath219 . in hamiltonian ( [ eq_full_version_of_persisten_flux_qubit_hamiltonian ] )
this changes @xmath220 the four tunneling operators are given by @xmath221 where @xmath222 corresponds to the junction with phase @xmath223 .
since the eigenstates are the same as for the p - flux qubit , it is straightforward to calculate the tunneling matrix elements .
the decay rates for quasiparticle tunneling through the large junctions are the same as for the standard persistent - current flux qubit and we therefore get the similar overall rates .
the only difference is that the tunneling across the small junctions , that form the squid , have now different type of interference effect if @xmath224 .
another version of the same qubit , the capacitively - shunted flux qubit @xcite has the same hamiltonian as the persistent - current flux qubit ( [ eq_full_version_of_persisten_flux_qubit_hamiltonian ] ) .
the major difference is that it is operated in the regime @xmath225 and the charging energy along the dimension of the double well potential is significantly reduced by introducing a shunt capacitance , @xmath226 .
this qubit has excellent coherence times and still preserves relatively large anharmonicity .
we will discuss the system at the symmetry point @xmath108 and for the case where it deviates most strongly from a harmonic oscillator , @xmath227 .
after this we are able to generalize this result to other parameter regimes . under the conditions we specified above we can write the hamiltonian as @xmath228 with @xmath229 .
we note that the hamiltonian is similar to the one of the strongly anharmonic phase qubit @xcite .
we find the eigenstates to be @xmath230 with @xmath231 .
it is now simple to calculate the matrix elements for the tunneling operators ( [ tunneling_operators_for_persitent_flux_qubit ] ) , @xmath232 and for @xmath233 one obtains again the additional phase factor @xmath234 .
the energy splitting between the eigenstates is given in a good approximation by @xmath235 .
using this result we can rewrite the matrix element as @xmath236 .
this is a well known result for a harmonic oscillator and is therefore also valid for the complete relevant parameter regime of the capacitively shunted flux qubit .
the interference angles @xmath137 of the three junctions are the same as for the p - flux qubit at the symmetry point : in the case of non - equilibrium quasiparticles and qubit decay the junctions 1 and 3 have constructive ( @xmath183 ) and the junction 2 has destructive interference ( @xmath237 ) .
vice versa for the dephasing .
as mentioned , the c - flux qubit is very similar to the strongly anharmonic phase qubit , with the difference that the latter qubit has only a single junction , which corresponds to the second junction of the c - flux qubit . as this junction shows destructive interference against qubit decay due to non - equilibrium quasiparticles , the strongly anharmonic phase qubit is better protected against this decoherence source .
however , the matrix elements of the qubits are generally similar to that of harmonic oscillators , and this means that the c - flux qubit is relatively well protected against quasiparticle tunneling , too . in table [ tab_comparing_decay_times ] we compare the qubit decay times between the capacitively shunted flux and the other discussed qubits .
in this work we analyzed decoherence in superconducting qubits due to quasiparticle tunneling .
we considered two types of sources of quasiparticle tunneling relevant for low temperatures and low qubit energies : non - equilibrium quasiparticles and sub - gap density of states .
using typically observed values for their densities we estimated the resulting qubit decay times and compared them with the experimentally measured ones . in the best case scenario , i.e. , with
the lowest reported quasiparticle densities , we showed that only the decay times of persistent - current flux qubits were similar to recent experiments .
the multi - junction flux qubits have achieved recently excellent coherence times and have the remarkable feature of extremely large anharmonicity which makes it the best approximation to a real two level system of all superconducting qubits .
however , the presented analysis shows that such qubits are also very fragile against quasiparticle tunneling induced decoherence . to protect the qubits against creation of non - equilibrium quasiparticles , for example , a careful isolation from the nearby environments
should be realized @xcite .
we thank a. heimes , w. d. oliver , p. kotetes , g. schn , j. liesenfeld and g. johansson for useful discussions .
this work was supported by the cfn of dfg and the u.s .
aro under contract no .
w911nf-09 - 1 - 0336 .
here we formulate the qubit decay rate as a function of electron correlation ( green s ) functions of the leads .
our aim is to consider the response of the superconductor as a sum over all energy states , rather than consider the response of single states . by this we can model situations where the superconductor quasiparticle density of states at a given energy can not be mapped back into certain quasiparticle energy in the normal state , for example , due to energy - level broadening effects .
we consider a single josephson junction qubit and write the time evolution of the reduced density matrix , when interacting with quasiparticle environment , generally as @xmath238 here the ( generalized transition rate ) tensor @xmath239 includes the effect of qubit - quasiparticle interaction .
its calculation is similar to the case of metallic reservoirs @xcite with a difference that when tracing out the environment also two annihilation ( creation ) operators @xmath240 can also contract to pairs .
their contribution leads to the pair density @xmath35 , whereas diagonal contributions , such as @xmath241 , lead to the density of states @xmath34 . in the leading order we obtain for the transition from the state @xmath242 to @xmath243 @xmath244
the trace over the initial distribution of quasiparticles leads to four nonvanishing contributions @xmath245 where @xmath246 and @xmath247 the correlation functions @xmath248 and @xmath249 are defined in eq .
( [ eq_greendefined ] ) .
we define their fourier transforms as @xmath250 the fourier tranforms are related to the spectral densities as in eq .
( [ eq_fouriertransforms ] ) . in our analysis
we have omitted the overall phase appearing usually in @xmath249-functions , as it is treated in the qubit part . in compared to refs .
our definition of @xmath251 corresponds to @xmath252 ( and @xmath253 to @xmath253 ) .
note the symmetry of the equation around @xmath257 : contribution of negative energies ( @xmath258 ) gives similar contribution as positive energies ( @xmath259 ) . as all processes described by @xmath260 correspond to given electron tunneling direction ,
negative energies correspond to opposite tunneling direction of the quasiparticle .
the contribution from @xmath261 follows the same calculation and is similar to @xmath260 , with a difference that the qubit matrix element is changed to @xmath262 .
it corresponds to tunneling of an electron to the opposite direction .
the contribution from @xmath263 and @xmath264 exists only if a factor @xmath265 is finite .
it means that it exists if two electron tunneling directions can lead to the same final state , in a single quasiparticle tunneling process .
then the processes interfere .
the contribution from @xmath263 is similar as before , but replaces @xmath34 functions by @xmath35 functions , where we have used the fact that for the considered qubits @xmath267 and then defined @xmath268 .
noticing , that the contribution from @xmath264 is equal to contribution from @xmath263 , except with an opposite phase factor @xmath269 , one obtains that the principal value part gives no contribution .
therefore one gets leading to equation ( [ eq_gamma12_defined ] ) .
this expression is similar to equation ( 9 ) in ref . ,
giving also insight to the interpretation of the different terms .
the main term that is missing , when compared to ref . , is the coherent josephson term ( @xmath272-term ) , as it does not contribute in the process considered here .
( this term is treated exactly in the qubit hamiltonian ) .
such coherent terms would contribute through the principal value integration .
the dissipative terms are similar .
especially , the famous @xmath71-term corresponds to interference between electron and hole - like tunneling in our calculation , described by terms @xmath263 and @xmath264 .
a new type of measurement of this term with a superconducting charge qubit has been considered in ref . .
taking into account the tunneling hamiltonian @xmath9 by perturbation theory includes not only incoherent processes but also coherent terms leading to renormalization of the qubit parameters .
their contribution is usually small but can depend , for example , on the non - equilibrium quasiparticle density .
this has been studied in detail in ref . .
the josephson coupling @xmath20 in eq .
( 2 ) is calculated as the expectation value over the quasiparticle distribution of the operator @xmath273 here the quasiparticle ( bcs ) hamiltonian has no dependence on the charge number @xmath11 .
using fermi distributions for the excitation occupation probabilities one obtains the famous ambegaokar - baratoff relation for @xmath20 .
this corresponds to the usual choice of the josephson coupling .
in addition , there exists a similar second - order contribution describing reactive behaviour of quasiparticles ( usually referred to as the capacitance renomalization ) @xmath274 to calculate this one can introduce a cut - off function @xmath275 $ ] , where @xmath276 is larger than any of the relevant energy scales in the system introduced to avoid divergence of the energy corrections . at this point
this term produces a constant energy shift for all @xmath11 and therefore does not bring anything new on the system .
insted of including these terms into the time - dependent problem , it is easier to estimate their effect by considering energy - level changes using time - independent perturbation theory . using the second - order theory again , but now for the modified tunneling hamiltonian @xmath9
, one obtains a correction to the energy of the state @xmath277 @xmath278\bar h_{\rm t}\vert i\rangle\right\rangle , \end{split}\ ] ] where @xmath279 is the energy of the qubit states @xmath242 . at this point
we see that , for example , the ambegaokar - baratoff result is a good approximation as long as the approximative josephson term @xmath280 and the corresponding term coming from the second term on the right - hand side of eq .
( [ correction ] ) almost cancel each other .
it occurs if the transition probabilities to virtual states @xmath281 ( @xmath282 ) are small , or if the energies of the virtual states stay small @xmath283 . in the opposite case extra anharmonicity ( non - constant energy - level spacing ) could emerge due to this correction , as the effective tunneling coupling can become different for different states .
the decoherence processes @xmath23 and @xmath24 correspond to leading order ( incoherent ) tunneling effects in the system .
the most important process in higher orders is usually the andreev tunneling .
this involves , for example , tunneling of two electrons from one side , with leaving two excitation behind , to form a cooper pair on the other side .
such processes dominate the sub - gap conductance of normal metal - insulator - superconductor junctions for ideal bcs state on the superconductor side .
however , for the case of perfect sis junctions no such process exist until the energy @xmath284 is somehow provided for creation of two excitations . the process appears as a step - like behaviour in the current - voltage characteristics of single jjs nearby @xmath285 . in experiments
the height of the step is usually considerably higher than obtained by theory . the reasons for this are still unclear @xcite . for the case of superconducting qubits
no energy is available to create two excitations , unless sub - gap density exists at one side of the junction . through creation of excitations into subgap region with creating a cooper - pair on the other side of the junction one
obtains a possible contribution to the decoherence .
however , a simple analysis indicates that the rates for such processes are proportional to @xmath286 , where @xmath287 is the effective number of parallel tunneling channels .
it results that contribution from this process should be much smaller than of the process @xmath24 considered in this paper .
we conclude that the contribution from higher - order tunneling effects to decoherence in superconducting qubits should stay small .
r. barends , j. wenner , m. lenander , y. chen , r. c. bialczak , j. kelly , e. lucero , p. omalley , m. mariantoni , d. sank , h. wang , t. c. white , y. yin , j. zhao , a. n. cleland , j. m. martinis and j. j. a. baselmans , arxiv:1105.4642 m. lenander , h. wang , radoslaw c. bialczak , erik lucero , matteo mariantoni , m. neeley , a. d. oconnell , d. sank , m. weides , j. wenner , t. yamamoto , y. yin , j. zhao , a. n. cleland and j. m. martinis , phys .
b * 84 * , 024501 ( 2011 ) .
m. w. johnson , m. h. s. amin , s. gildert , t. lanting , f. hamze , n. dickson , r. harris , a. j. berkley , j. johansson , p. bunyk , e. m. chapple , c. enderud , j. p. hilton , k. karimi , e. ladizinsky , n. ladizinsky , t. oh , i. perminov , c. rich , m. c. thom , e. tolkacheva , c. j. s. truncik , s. uchaikin , j. wang , b. wilson and g. rose , nature * 473 * , 194 ( 2011 ) .
j. majer , j. m. chow , j. m. gambetta , j. koch , b. r. johnson , j. a. schreier , l. frunzio , d. i. schuster , a. a. houck , a. wallraff , a. blais , m. h. devoret , s. m. girvin , and r. j. schoelkopf , nature * 449 * , 443 ( 2007 ) . | we study decoherence in superconducting qubits due to quasiparticle tunneling which is enhanced by two known deviations from the equilibrium bcs theory .
the first process corresponds to tunneling of an already existing quasiparticle across the junction .
the quasiparticle density is increased , e.g. , because of an effective quasiparticle doping of the system .
the second process is quasiparticle tunneling by breaking of a cooper pair .
this can happen at typical energies of superconducting qubits if there is an extended quasiparticle density inside the gap .
we calculate the induced energy decay and pure dephasing rates in typical qubit designs . assuming the lowest reported value of the non - equilibrium quasiparticle density in aluminum
, we find for the persistent - current flux qubit decay times of the order of recent measurements .
using the typical sub - gap density of states in niobium we also reproduce observed decay times in the corresponding niobium flux qubits . |
atom optics has obtained an unprecedented development since the advent of bose - einstein condensates ( bec s ) of dilute alkali atomic gases @xcite . for instance , the applications of atom optics have been realized such as atom laser , solition generation @xcite and matter wave interferometry @xcite . besides ,
many new techniques of manipulating the cold atoms have been devised such as bragg scattering @xcite and magnetic waveguide @xcite , etc .
these sophisticated techniques may lead to reaching the goal in higher precision measurement @xcite and applications in quantum information processing @xcite .
recently , saba _ et .
@xcite and shin _ et .
@xcite have demonstrated that the josephson effect by using the technique of atom optics , bragg scattering .
two beams of atoms are optically extracted from two separate trapped bose - einstein condensates ( bec s ) using bragg scattering @xcite .
then , these two beams of atoms overlap and interfere with each other and the measurement process creates a relative phase between the two bec s .
they showed that the josephson coupling of these two spatially separate systems can be made through an intermediate coupling system .
it is quite different to the conventional josephson devices , such as superconducting systems @xcite , and bose - einstein condensates @xcite , in which the two quantum systems are spatially connected to each other and the wavefunction of them have small overlap . in this paper
, we investigate that the two trapped condensates are connected by outcoupling a small fraction of condensates via the bragg scattering @xcite and those outcoupling atoms are transported through a magnetic waveguide @xcite to replenish the trapped condensates @xcite .
we show that this system is equivalent to the usual hamiltonian describing the josephson effect of bec s trapped in a double - well potential .
the strength of the josephson coupling can be explicitly controlled by varying the outcoupling rate of atoms using the bragg beams whereas the coupling phase between these two condensates can be tuned by adjusting the phase shifts of the outcoupled atoms @xcite . moreover ,
the spontaneous scattering due to the bragg beams is negligible by virtue of the large detuning of the pulses .
this `` optical josephson junction '' could be used , for example , to implement the precision scheme proposed by dunningham and burnett @xcite in which they consider the precision measurement using an entangled bec s trapped in a double - well potential .
we now study the implementation of the precision measurement scheme being suggested by dunningham _
@xcite , which can measure the nonlinear interaction strength and gravity , using this `` optical josephson junction '' .
according to this proposal , the measurement of phase can be done by appropriately turning on and off the josephson coupling .
then , the final phase information is encoded in the number fluctuation that can be detected from the collapse and revival of the relative phase between the condensates .
this scheme provides a simple way to measure phase with heisenberg limited accuracy , i.e. , the phase uncertainty scales as @xmath0 , where @xmath1 is the number of atoms .
the schematic of the optical coupling of two trapped condensates 1 and 2 , and the outcoupling atoms are transferred via a waveguide in the ring form .
the pair of bragg beams with frequencies @xmath2 and @xmath3 are denoted by dash - dot line and dash line respectively.,height=226 ] the paper is organized as follows : in sec .
ii , we introduce the model of this `` optical josephson junction '' . in sec .
iii , we derive an effective hamiltonian under the two - mode approximation of two trapped bec s . in sec .
iv , we study the implementation of dunningham _ et al .
_ scheme in this system .
we briefly review the mechanism of the bragg scattering in appendix a.
we consider two trapped condensates connected to a 1d ring - form waveguide using the @xmath4-th order bragg scattering @xcite as shown in fig 1 .
in fact , this bragg scattering is a @xmath5-th order multi - photon raman process @xcite .
the stimulated emission and stimulated absorption are made by using the two bragg beams being acted as a pump and probe field with counter - propagation and slightly frequency difference , i.e. , @xmath6 and @xmath3 .
the wavelength and wave number of the bragg beam with frequency @xmath2 are @xmath7 and @xmath8 respectively .
this pair of bragg beams give out a certain amount of momentum to the two trapped bec s @xmath9 and @xmath10 respectively .
then , a small amount of bec s are outcoupled from these two trapped bec s with definite momenta @xmath11 and @xmath12 at the @xmath4-th order raman process . without loss of generality , we assume the two momenta @xmath13 and @xmath14 with the equal magnitude but opposite sign , for @xmath15 and @xmath16 is the mass of an atom .
these outcoupling condensates are then transported via the waveguide to the other trapped bec .
this conservation of total number of particles may enable us to overcome the limitation of the coupling time in the experiments @xcite in which the atoms were linearly coupled out and not transferred back to the system @xcite .
but now the coupling process can be continued if the number of atoms are conserved in the system .
we discuss the hamiltonian of this system and derive an effective two - state hamiltonian to describe the josephson effect .
the hamiltonian of the total system has the form @xmath17 where @xmath18 , @xmath19 and @xmath20 are the hamiltonians of ; bec s 1 and 2 , the outcoupled atoms in the 1d ring , and the coupling between the trapped atoms and the ring respectively .
the hamiltonian for the trapped atoms has the form and can be expressed in terms of the arc length , @xmath21 for @xmath22 is the radius of the ring and @xmath23 is the polar angle of the ring , @xmath24 \psi_j(s)\nonumber\\ & & + \frac{u_0}{2}\psi^\dag_j(s)\psi^\dag_j(s)\psi_j(s)\psi_j(s)\bigg\},\end{aligned}\ ] ] where @xmath25 and @xmath26 are the field operator and trapping potential respectively for condensate @xmath27 , @xmath28 is the interaction strength and @xmath29 .
we consider that the two condensates are held in two different traps that are sufficiently deep , and hence the single - mode approximation can be applied @xcite .
this allows us to describe the trapped condensates as the local ground state of the trap .
the field operator @xmath30 can then be approximated by @xmath31 , where @xmath32 and @xmath33 are the annihilation operators and the mode function of the potential of the @xmath27-th trap and @xmath29 .
if we take the system to be symmetric so that @xmath34 has the same form as @xmath35 , the hamiltonian @xmath18 can be written as @xmath36,\end{aligned}\ ] ] where @xmath37\psi_j(s),\end{aligned}\ ] ] is the eigenenergy of the two modes and @xmath38 is the self - interaction strength .
the free condensates can be outcoupled from the trapped condensate using the bragg scattering @xcite .
we briefly review its mechanism in the appendix a. the nearly perfect efficiency of the first order bragg resonance has been observed experimentally @xcite . however , the efficiency of the high order bragg scattering is decreased significantly @xcite .
the hamiltonian representing the effective coupling between the trapped atoms and the free condensate is written , thus , @xmath39+{\rm h.c.},\ ] ] where @xmath40 is the outcoupling amplitude between two trapped bec s and free outcoupling condensates @xcite , @xmath41 @xciterespectively . here
, we can consider the coupling between the trapped condensates and the free condensates with momenta @xmath42 .
moreover , we consider the replenishment process in which the free condensates are deaccelerated by the bragg beams and merge into the trapped condensates @xcite .
we assume that the excitation in this replenishing process is negligible . on the other hand ,
the relative phase shift @xmath43 is generated during the flight between the two condensates .
the phase shift @xmath44 directly depends on the bragg beams and the momenta @xmath42 which can be adjusted in the experiment @xcite .
the hamiltonian of the 1d ring has the corresponding form : @xmath45,\end{aligned}\ ] ] where @xmath46 is the field operator of the condensate in this ring - form waveguide transported by a waveguide . we approximate these unconfined in the angular direction , but confined radially and weakly interacting condensate as the freely propagating non - interacting condensates
. we can justify this approximation by considering the magnitude of the kinetic energy and the strength of the nonlinear interaction .
the kinetic energy @xmath47 of the outcoupled condensates by the bragg beams is with @xmath48 khz @xcite whereas the nonlinear interaction of this nearly free condensates @xmath49 is about @xmath50 khz @xcite .
here @xmath51 is density of the outcoupled condensate which must satisfy @xmath52 and @xmath53 is the scattering length of the atom . from the above estimation
, we can argue that this weakly interacting outcoupled condensates as free condensates .
the condensates in the ring are composed of two free condensates outcoupled from the two trapped condensates as @xmath54 where @xmath55 is the circumference of the loop , @xmath42 should satisfy the boundary condition @xmath56 and @xmath57 is an integer .
then , the hamiltonian of the ring can be written as @xmath58 where @xmath59 .
we now derive an effective hamiltonian to describe the josephson effect between two trapped bec s .
the free propagating condensates act as immediate states to couple the two trapped condensates .
the coupling hamiltonian is given by @xmath60 where @xmath61 , @xmath29 and @xmath62 .
we assume that the two coupling strengths @xmath63 are similar by letting @xmath64 .
the heisenberg equation of motion of @xmath65 is @xmath66 we use the adiabatic approximation , i.e. , @xmath67 as the condition @xmath68 is satisfied : @xmath69 we can see that it is true if we can assure that @xmath70 is much less than @xmath71 . physically speaking
, the validity of this adiabatic approximation assumes that the transition time between these intermediate states and the two trapped states is short enough .
moreover , it is noted that the whole @xmath5-th order raman process requires a finite time duration with the interaction of laser to be completed .
this time duration must be short compared to the time of the josephson coupling .
the resulting effective two - state hamiltonian can thus be written as @xmath72\nonumber\\ & & -\frac{{\hbar}g}{2}\big(e^{-i\theta}c^\dag_1c_2+e^{i\theta}c^\dag_2c_1\big),\end{aligned}\ ] ] where @xmath73 .
it is noteworthy that this effective two - mode hamiltonian is akin to the josephson hamiltonian of the external @xcite and the internal @xcite bec systems .
this optical josephson coupling can then yield the josephson effect of two condensates .
likewise , this josephson coupling can be controlled by the strength of the bragg beam to vary @xmath74 whereas the phase of josephson coupling can be adjusted by the phase shift @xmath75 that can be done by changing @xmath76 and the arc length between the bec s .
indeed , this feature is very useful to the application of this josephson junction .
it is convenient to write the effective hamiltonian in terms of the angular momentum operators as @xmath77 where the total atom numbers @xmath78 is conserved .
then , the hamiltonian is written as @xmath79 where @xmath80 .
this constant @xmath81 does not affect the quantum dynamics of the system .
therefore , we will ignore it in the subsequent discussion .
in this section , we discuss the implementation of precision measurement on the optical josephson junction .
dunningham and burnett proposed a precision measurement scheme using a number - squeezed bec trapped in a double - well potential @xcite .
this scheme requires the active control of the josephson coupling .
the phase information can be finally obtained by measuring the number fluctuations which can be detected via the visibility of the interference fringes of the two condensates . however , there are some limitations of the bec double - well system with respect to the dynamical control of the josephson coupling .
the josephson coupling strength exponentially depends on the height of potential barrier between the two wells .
the potential barrier has to be tuned very accurately in order to vary the coupling . besides , when the josephson coupling is large , there needs to be a large spatial overlap between the two localized wave functions in two wells . in this case , the usual two - mode approximation is no longer valid .
this problem can now be avoided in the optical josephson junction system due to the large separation between the two trapped condensates . since the optical josephson junction is a well - controlled system , it is a promising model to implement this scheme .
we proceed to describe how to realize the protocol of the scheme proposed by dunningham _
et al._. if we choose the phase shift @xmath82 , the effective hamiltonian has the form , @xmath83 we first prepare the initial state as a number - squeezed state @xmath84 which contains a definite number of atoms trapped in two wells respectively .
this can be prepared by adiabatically switching off the josephson coupling strength so that the condensates are isolated in two different traps in the fock regime , i.e. , @xmath85 @xcite .
we can use this system to measure a relative phase between the condensates as follows .
a large josephson coupling is turned on rapidly with the strength @xmath86 by using bragg pulses with @xmath87 .
it is convenient to write the state in the new eigenbasis , i.e. , the symmetric mode @xmath88 and antisymmetric mode @xmath89 with respect to two traps .
the quantum state is in this new basis , has the form @xcite @xmath90 .
it is worth noting that this superposition of states is robust against the particle loss @xcite . in this regime
, there is a small energy difference , @xmath91 , between the symmetric and antisymmetric modes , different phases result for the terms in the superposition . then , the system is held for a certain time @xmath92 to allow the natural evolution of the system .
next , the josephson coupling is suddenly switched off fast enough with respect to coupling between wells , but slow with respect to energy level spacing in each well .
thus , the state is conveniently expressed in terms of the number basis in each well . the quantum state becomes @xcite @xmath93 for @xmath94 and @xmath95 .
this completes the measurement of the phase @xmath23 which is recorded in the quantum state now .
the information of this phase @xmath23 can be obtained from the number uncertainty @xmath96 , where @xmath97 is the number of atoms in the trap 1 ( or 2 ) . from eq .
( [ state_phi ] ) , the number uncertainty can be calculated as @xcite @xmath98 this number uncertainty is of order @xmath1 . according to the number - phase uncertainty relation @xmath99 , we can see that the uncertainty of the phase @xmath23 , for the minimum uncertainty state , is of order @xmath100 , i.e. , heisenberg limited .
this number variance can be experimentally determined from the interference pattern .
bragg scattering provides a non - destructive method to determine the relative phase . a small fraction of condensates
are coupled out horizontally from these two condensates and allowed to interfere with each other .
the relative phase between the two trapped condensates can be determined from the interference patten of these two outcoupled condensates @xcite .
the interference pattern of the fringes between the two outcoupled condensates @xcite , @xmath101 $ ] , is directly proportional to the interference terms of two trapped condensates , where @xmath102 and @xmath103 are the field operator of two outcoupled condensates from the @xmath27-th trap , and @xmath104 is coordinate in the horizontal direction . following the similar treatment to the previous section and taking @xmath82 ,
the intensity of the interference fringes , @xmath105 , of these two outcoupled condensates can be obtained as @xcite @xmath106 where @xmath107 is the time of holding the system with the nonlinear interaction of the atoms .
this state vector is given by @xmath108\tau/2}d_n|n\rangle_1|n - n\rangle_2.\ ] ] thus , the intensity , @xmath105 , is proportional to @xmath109 the collapse times @xmath110 can be estimated by considering the particle numbers in the range , @xmath111 .
hence , the collapse times of the relative phase @xmath110 is about @xmath112 @xcite .
this collapse time can be determined by holding the system with the nonlinear interaction as a function of time @xmath107 and measure the corresponding intensity @xmath105 with different @xmath107 s .
therefore , we can determine the number fluctuation and hence the required phase information . from an experimental point of view
, the collapse time is very short and so the damping and decoherence effects may be minimized .
hence , the number variance may be able to be measured accurately by this method .
although the revival time can reveal the phase information , it takes a much longer time to observe . bear in mind that the measurement of collapse and revival time of a bose - einstein condensate has been demonstrated in the experiment @xcite .
it is noted that the low efficiency of the higher order bragg scattering may limit the effective coupling of the two coupled bec s .
nevertheless , we can still justify the validity of this optical based josephson junction as discussed above . on the other hand , it is very interesting to compare this optical based josephson junction to the solid - state counterpart . in summary
, we theoretically study the microscopic model of the `` optical josephson junction '' and derive the effective hamiltonian of this device .
we have discussed how this system can be used to implement hesienberg limited precision measurement .
this measurement can be made by detecting the collapse time non - destructively .
h.t.n . was supported by the croucher foundation .
thanks the royal society and wolfson foundation for support .
in this appendix , the basic mechanism of the condensates using the bragg scattering is briefly reviewed .
the pump - probe mechanism has been discussed in detail in the reference @xcite .
the pump and probe fields impart a momentum @xmath113 to the ground state of the condensates each time by coupling to the excited state with a large detuning @xmath114 between the two - level atoms with the energy difference @xmath115 and the laser field . to elucidate this process , we study the first order bragg resonance case by considering a time - independent hamiltonian in the interaction picture and assume the zero ground state energy for @xmath116 which is given by @xmath117 where @xmath118 and @xmath119 are the annihilation operators for the excited and ground states with the momentum @xmath120 , @xmath121 and @xmath122 are the coupling strength of the pump field and the probe field of this pair of bragg beams ; and @xmath123 and @xmath124 are the detuning between the ground and the excited states with the different momenta and the bragg beams , for @xmath125 .
the equations of motion for the different momentum states are given by : @xmath126 at the bragg resonance , the detuning @xmath127 equals zero at @xmath128 .
the excited state @xmath129 can be adiabatically eliminated as @xmath130,@xmath122 , i.e. , @xmath131 therefore , the equations of motion for these two ground states with different moment have the form : @xmath132 clearly , we can see that these two momentum states are effectively coupled with each other at the first order bragg resonance . in general
, we can consider the @xmath4-th order bragg scattering which is a @xmath5-th multi - photon raman process . within this process ,
the different momentum modes are virtually excited but they can be adiabatically eliminated because of energy conservation being unfavourable .
it is legitimate to consider the effective coupling between the trapped condensates and the free momentum states @xmath133 at the bragg resonance only in which the energy is conserved .
the explicit form of the effective coupling between the trapped and free states , @xmath40 , has been found as @xcite : @xmath134 ^ 2\omega^{m-1}_{2k } } \bigg|.\end{aligned}\ ] ] the detailed analysis can be referred to the reference @xcite . | we study a new type of josephson device , the so - called `` optical josephson junction '' as proposed in phys .
rev .
lett . * 95 * , 170402 ( 2005 ) .
two condensates are optically coupled through a waveguide by a pair of bragg beams .
this optical josephson junction is analogous to the usual josephson junction of two condensates weakly coupled via tunneling .
we discuss the use of this optical josephson junction , for making precision measurements . |
in many - body physics , the correlation function plays a fundamental role . especially in theoretical studies , the understanding on many physical phenomena are based on the calculation of the corresponding correlation functions@xcite . for instance , to investigate the magnetic properties of a system , people often calculate the spin - spin correlation function to learn the possible magnetic order . in 1d case ,
the system has a ferromagnetic order if the @xmath0 mode is dominant , and an anti - ferromagnetic order if the @xmath1 mode is dominant .
the long - range behavior of the correlation function is associated with the symmetry - breaking in the system , which is an important concept in the understanding of continuous phase transitions .
that is , for a certain operator , its non - vanishing value at a long distance denotes a symmetry - broken phase which usually occurs in the thermodynamic limit .
mathematically , if the correlation function @xmath2 is a constant in the infinite @xmath3 limit , we say that the state has a long - range order .
the operator @xmath4 can then be taken as the order parameter to describe the corresponding phase . traditionally , to find an appropriate order parameter for a certain phase
, people often resort to other methods such as group theory and the renormalization group analysis .
nevertheless , these methods can not always help us to find the correct form of the order parameter .
hence a method to derive the order parameter systematically is very instructive . using the variational approach , furukawa _ _
et al__@xcite proposed a scheme to derive the order parameter .
while the scheme is promising , it still needs the knowledge of the degenerated states that lead to the symmetry breaking in the thermodynamic limit .
instead , we proposed recently a scheme to derive the order parameter from the spectrum of the reduced density matrix of the ground state directly @xcite .
differ from furukawa s scheme , our approach is not variational , and needs only the knowledge of the ground state . in this paper
, we apply the scheme , for the first time , to a realistic model . to start with
, we will first demonstrate the derivation of the order parameter in the hubbard model in section [ sec : hm ] .
then , we apply our scheme to the extended hubbard model ( ehm ) in section [ sec : ehm ] and [ sec : ehm2 ] .
we show that even without any knowledge of the symmetry of the system , we can derive the exact forms of the order parameters which show long - range correlations in the ground state of the model .
a summary would be given in section [ sec : sum ] . and
@xmath5 in the hubbard model .
the inset shows that the mutual information decays algebraically with @xmath3 and thus long - range correlation is presented in the system.,width=306 ]
the hubbard type models has been served as a prototype in condensed matter physics to study the electron correlation effect in solids . especially in many cases where the long - range coulomb interaction has significant importance , like in quasi one - dimensional organic solids such as conjugated polymers @xcite and charge transfer crystals @xcite
, the extended hubbard model in one - dimension is the simplest model that included finite range interactions for theoretical studies . the model s
hamiltonian reads@xmath6where @xmath7 and @xmath8 ( @xmath9 ) are creation and annihilation operators of electrons with a spin @xmath10 at the site @xmath11 respectively , @xmath12 , and @xmath13 .
@xmath5 and @xmath14 is the strength of the on - site and the nearest - neighbor coulomb interaction respectively .
@xmath15 is the hoping integral and is taken to be unity for convenience .
interestingly , the ehm exhibits a very rich phase diagram .
it was shown analytically @xcite and numerically @xcite the existence of spin - density waves ( sdw ) , charge - density waves ( cdw ) , phase separation ( ps ) , singlet superconducting ( ss ) and the triplet superconducting ( ts ) phases in the model . using concepts from quantum information theory as a tool , it is also shown that the local entanglement @xcite was able to obtain the cdw , sdw and ps phases .
in addition to these three phases , the block - block entanglement can also witness the ss phase @xcite .
besides the phases mentioned above , there are also studies pointed out the existence of the bond - order waves phase(bow ) in the model .
however , whether this bow exists as a narrow region or just a line in the ground state phase diagram is still a controversial problem @xcite . at this point ,
let us first forget about the result from previous studies and suppose we know nothing about the nature of the phases . to find the potential order existing in the ground state of the model , we need to first examine if the ground state exist a long - range correlation or not .
while we do not know the form of any order parameters , we can still use the mutual information , a concept borrowed from quantum information science , to measure the total correlation between two arbitrary blocks in the system .
the mutual information is defined as@xmath16where @xmath17tr@xmath18 is the von - neumann entropy of the block @xmath19 .
@xmath20 is the reduced density matrix obtain by tracing out all other degrees of freedom except that of site @xmath19 , i.e.@xmath21tr@xmath22 where @xmath23 is the ground state of the system .
it has been shown that the existence of the long - range order means that there is a non - vanishing mutual information at a long distance @xcite .
this statement still holds true vice versa . by studying the spectra of the reduced density matrices
, we can derive the potential order parameters . to illustrate the scheme more explicitly ,
let us first ignore the next - nearest neighbor interaction , i.e. @xmath24 , and the model is reduced to the conventional hubbard model @xcite .
consider a block consists of one site , i.e. @xmath25 . in the basis of local states
@xmath26 , @xmath27 is found to be diagonal , i.e. @xmath28 where @xmath29 and @xmath30 are some positive real numbers .
the two - site reduced density matrix @xmath31 is a block - diagonal matrix in the basis @xmath32 .
figure 1 shows the dependence of the mutual information @xmath33 as a function of @xmath34 and @xmath5 calculated for a 14-site system with 7 spin - up electrons and 7 spin - down electrons .
the results are obtained from numerical exact diagonalization under periodic boundary conditions . from the figure
, we can see that the mutual information reaches a maximum at @xmath35 , which is a non - trivial point of the hubbard model .
we will return to this phenomenon later . at this stage
we are more interested in the long - range behavior of the mutual information . for this purpose , we plot the dependence of @xmath33 on @xmath3 in the inset of figure 1 for various @xmath5 . from the inset , it is easy to find that @xmath36 decays algebraically with @xmath3 .
therefore , we judge that in the ground state of the 1d hubbard model , there must exist certain kind of long - range correlation though we do not know the explicit form of the order parameter yet . as a function of @xmath3 for two limiting cases of @xmath37 in the hubbard model . for @xmath38 ( left ) , the charge degrees of freedom ( @xmath39 ) are correlated , while for @xmath40(right ) the spin degrees of freedom(@xmath41 ) are correlated . ,
width=340 ] as a function of @xmath5 in the hubbard model . here
@xmath42 and @xmath25 .
for the negative @xmath5 case , the eigenstates @xmath43 and @xmath44 have dominate weights while for positive @xmath5 , the eigenstates @xmath45 and @xmath46 have dominate weights.,width=306 ] _ diagonal order parameter : _ according to the scheme we proposed recently@xcite , the order parameter can be constructed from the spectrum of the reduced density matrix @xmath47 .
however , in most of the cases , we can not obtain results of the total correlation between two blocks separated by a large enough distance due to the limitation of computational powers . to single out the correlated elements in @xmath48 , we calculated the difference between the diagonal matrix elements of @xmath49 and that of @xmath50 as a function of the separation @xmath51 .
figure [ figure_ev_rdm1dddd ] shows a plot of @xmath52 as a function of @xmath3 . for convenience ,
the two block reduced density matrix @xmath53 is written as @xmath54 . from the figure , clearly we can see that @xmath55 which is consistence with @xmath56 .
moreover , for @xmath38 , the difference in the diagonal matrix elements for @xmath57 decays algebraically with @xmath3 while it is almost zero for the others .
we can then argue that the vacant state and the double occupancy state of the two blocks are correlated in this case .
for the case of @xmath40 , @xmath58 are almost zero unless @xmath59 . in this limit , the spin up and spin down states are correlated . in figure [ figure_ev_rdm1 ]
, we show the dependence of the eigenvalues of @xmath48 on the interaction @xmath5 . from the figure
, we can see that if @xmath5 becomes larger and larger , the eigenvalues of the state @xmath60 and @xmath61 tend to 1/2 and the eigenvalues of @xmath62 and @xmath63 vanish .
the observation means that in the large @xmath5 limit , the reduced density matrix becomes @xmath64 in the basis of @xmath65 , we show the non - zero diagonal and off - diagonal elements of @xmath31 in figure [ figure_rdm2_r7_dia ] and figure [ figure_rdm2_r7_offdia ] respectively . from the figures , we find that @xmath66 takes the form @xmath67 in the large @xmath5 limit in the basis of @xmath68
. therefore , the order parameter can be written as @xmath69according to the traceless condition tr@xmath70 , we have @xmath71 .
let @xmath72 , then the order parameter becomes@xmath73for simplicity , we denote it as @xmath74 .
similarly , for the case of negative @xmath5 , we can find that @xmath75and @xmath76 in the basis of @xmath77 .
then the diagonal order parameter can be defined as@xmath78apply the traceless condition and cut - off condition , we get @xmath79for simplicity , we denote it as @xmath80 . as a function of @xmath5 in the hubbard model for @xmath81 . in the negative @xmath5 limit
, only the matrix elements corresponds to the basis @xmath82 is non - vanishing . in the positive @xmath5 limit
, only the matrix elements corresponds to the basis @xmath83 is non - vanishing.,width=321 ] as a function of @xmath5 in the hubbard model for @xmath84 .
the matrix elements for @xmath85 is the same as @xmath86.,width=321 ] _ off - diagonal order parameter : _ from eq .
( [ eq : singlerdm ] ) , we find that @xmath66 is not a diagonal matrix .
this observation means that there exists off - diagonal long - range correlations in the ground state of the 1d hubbard model . according to our scheme
, the off - diagonal order parameter can be defined as @xmath87here @xmath88 is complex , so there are two independent parameters in the operator .
we separate the real part and the imaginary part in the operator @xmath89with @xmath90obviously , we have @xmath91 and @xmath92 so we can treat either @xmath93 or @xmath94 or their linear combination as the off - diagonal order operator .
let s denote them as @xmath95 and @xmath96 respectively .
similarly , in the negative @xmath5 region , we can also derive the off - diagonal order parameter as@xmath97we denote them as @xmath98 and @xmath99 respectively .
@xmath100 _ algebra and the effective hamiltonian : _ in the basis of @xmath101 , @xmath102 and @xmath74 can be written as@xmath103these expressions are exactly the form the pauli matraces , which satisfy the su(2 ) lie algebra , i.e. @xmath104=2i\delta _ { ij}\epsilon ^{\alpha \beta \gamma } \sigma _ { i}^{\gamma } $ ] . as we know , in the large @xmath5 limit , the hubbard model becomes the one - dimensional heisenberg model @xmath105where @xmath106 . in the ground state of the heisenberg model ,
the spin - spin correlation has a power - law decaying behavior .
the order parameters derived is consistent with our knowledge .
on the other hand , in the negative @xmath5 region , @xmath98 , @xmath107 , and @xmath80 can be written as@xmath108 in the basis of @xmath109 .
clearly the three operators satisfy the su(2 ) lie algebra too . for @xmath110 in ehm .
the inset shows a ln - ln plot of the mutual information versus @xmath3 for @xmath111.,width=306 ] as a function of @xmath3 for @xmath112 ( top panel ) , @xmath113 ( middle panel ) , and @xmath114 ( bottom panel ) in ehm for @xmath110 . for @xmath115 ,
the charge degrees of freedom ( @xmath116 ) are correlated , while for @xmath113 the spin degrees of freedom(@xmath117 ) are correlated.,width=306 ] now , let us also include the next - nearest neighbor interaction in the system .
figure [ figure_entro_ehm ] shows a plot of the mutual information as a function of @xmath3 .
obviously , the mutual information is non - vanishing at a large @xmath3 for @xmath119 ( in fact , they corresponds to two different regime where @xmath120 and @xmath121 which we will see more explicitly later ) . for the intermediate case
, we can see from the log - log plot in the inset that the mutual information shows a power law decaying behavior .
one can then safely argue that the system exhibits certain kind of long - range correlations and we can go on to investigate the spectrum of the reduce density matrix to derive the potential order parameters . using the same basis as that for the hubbard model in the previous subsection
, we calculated the single - site and two - sites reduced density matrices .
@xmath20 is diagonal and takes the form of eq .
[ eq : rhoi ] .
figure [ figure_diff_ehm ] shows a plot of the difference between the diagonal matrix element of @xmath54 and the product of the diagonal matrix elements of @xmath122 and @xmath123 with the corresponding basis as a function of @xmath3 .
the finite difference in @xmath124 for some particular @xmath125 and @xmath126 indicates that @xmath127 . for @xmath112 and @xmath114
, we see that the main contributors to the correlation of two sites separated at a large distance are the states @xmath43 and @xmath44 . while for the case of @xmath113 , the @xmath45 and @xmath46 states play the role .
figure [ figure_ev1_ehm ] plots the eigenvalues of the single - site reduced density matrix as a function of @xmath14 .
the crossings of the eigenvalues , which is the probabilistic weight of the corresponding eigenstates , of @xmath20 separate the system into three different regimes corresponding to @xmath128 , @xmath129 , and @xmath120 .
this echoes what we have mentioned in the previous paragraph . in each of the regime ,
the kind of correlation existing in the system is qualitatively different as one can judge from the nature of the dominating eigenstates . for @xmath129 ,
the eigenvalues of the @xmath46 and @xmath45 states are around @xmath130 while that are almost zero for the @xmath43 and @xmath44 states . following similar argument in the previous section
, we can define the order parameter as @xmath131 again , by applying the traceless condition and take @xmath132 , we obtain the order parameter @xmath133 for @xmath128 and @xmath120 , the eigenvalues corresponds to the states @xmath43 and @xmath134 are around one - half while that for the states @xmath45 and @xmath46 tends to zero .
note that the discrepancy from the value @xmath130 and @xmath0 respectively for @xmath128 is due to the finite size effect .
nevertheless , we show the size dependence of @xmath135 for @xmath136 in the inset of fig .
[ figure_ev1_ehm ] .
the eigenvalues for the @xmath45 and @xmath46 becomes vanishing as the system size increases .
we can define the order parameter as @xmath137 the choice of the order parameter can be narrow down by applying the traceless condition and again taking @xmath138 to be @xmath139 .
we have @xmath140 as a function of @xmath14 for @xmath110 in ehm .
the inset shows the size dependence of the eigenvalues of @xmath20 for @xmath136 . for @xmath128 and @xmath120 , the eigenstates
@xmath43 and @xmath44 have dominate weights . for @xmath141 , the eigenstates @xmath45 , and @xmath46 are dominated.,width=321 ] in ehm .
here @xmath110 , @xmath42 , and @xmath112 ( top panel ) , @xmath113 ( middle panel ) , and @xmath114 ( bottom panel ) respectively.,width=302 ] here we obtained the same order parameter for both of the regimes where @xmath128 and @xmath120 . however , recall from fig .
[ figure_entro_ehm ] , the mutual information as a function of @xmath3 shows qualitatively different behavior in these two range of values of @xmath14 . for @xmath120 , the mutual information almost remains constant as @xmath3 increases . on the other hand , for @xmath128 ,
the mutual information first decreases and then increases to a maximum value upon reaching @xmath84 , where it is the middle of the chain .
it is reasonable to suspect that the ground state of the system is qualitatively different in these two regimes . to further distinguish between them
, it is worth to study the mode of the order parameters . with the order parameter found in eq .
( [ eq : o_cdw ] ) , one can calculate the correlation function in eq .
( [ eq : corr ] ) and it is found to be oscillating with respect to @xmath3 as expected . to capture the mode of the order parameter ,
let s consider the fourier transformation of the correlation function , i.e. @xmath142 where @xmath143 and @xmath144 .
also note that @xmath145 as a result of applying the traceless condition .
the result is plotted in fig .
[ figure_cor_ehm ] . from the figure
, we see that the correlation function peaks at @xmath146 for @xmath114 .
one can expect this holds true for the whole range of @xmath120 .
together with the form of the derived order parameter , we may conclude that the dominating configuration in the ground state of the system are consisting of alternating vacant and double occupancy states .
this is in fact the well - know charge - density wave ( cdw ) states @xmath147 and @xmath148 in the extended hubbard model . for @xmath128 ,
the mode of the order parameter are @xmath149 and @xmath150 .
the second peak in the correlation function @xmath151 is just a result from the periodic boundary conditions . in this regime
, one can deduce that the dominating ground state configurations has a period of half of the lattice in the real space .
they are the phase separation ( ps ) states @xmath152 and the translation of it .
this also explains why the mutual information is maximum for the sites separated by half of the lattice . since any of the translation of the above state are equally weighted in the ground state before symmetry
is broken , only the local states separated by half of the lattice can be confidently determine once we know one of them .
they have to be opposite to each other . for completeness
, we also studied the mode of the order parameter for @xmath141 .
the correlation function @xmath153 as a function of @xmath154 is shown in the middle panel of fig .
[ figure_cor_ehm ] .
the maximum of the correlation function occurs at @xmath146 .
we can similarly deduce that the dominating configuration are the spin - density wave ( sdw ) states @xmath155 and @xmath156 .
these results obtained are consistent with previous studies @xcite .
let s now consider a block size of two sites , i.e. @xmath159 . in the following , we will use @xmath19 to denote a single - site and @xmath160 to denote a single - block consist of two neighboring sites .
the single - block ( two - site ) reduced density matrix @xmath161 is calculated for the case of @xmath158 and @xmath42 . in the basis of @xmath162 ,
the eigenstates of @xmath161 have the form of @xmath163 where @xmath164 are positive real numbers . as a function of @xmath14 in the 1d extended hubbard model . @xmath42 and @xmath158 .
the inset shows a close - up of the low - lying states.,width=321 ] and @xmath165 in the eigenstate @xmath166 of the single - block reduced density matrix as a function of @xmath14 .
@xmath42 , @xmath158 , and @xmath159.,width=302 ] figure [ fig : rdm2_ehm ] shows a plot of the eigenvalues of the corresponding eigenstates as a function of @xmath14 . from the figure
, we can see that some of the states , for examples , states @xmath167 and states @xmath168 , are degenerated respectively .
the degeneracy tells us that there are spin - up - down and charge symmetries in the system . among all the eigenstates ,
the weight of the state @xmath166 is dominated for the whole range of value of @xmath14 shown . this state is a spin - singlet state .
it is a superposition of two parts , the charge part @xmath169 and the spin part @xmath170 with relative weight characterized by @xmath171 and @xmath165 respectively .
the magnitude of @xmath171 and @xmath165 are plotted as a function of @xmath14 in fig .
[ fig : rdm_p_ehm ] .
there is a crossing between the two magnitudes around @xmath172 . for @xmath173 ,
the relative weight of the charge part is much greater than that of the spin part . besides , considering the region @xmath173 in fig .
[ fig : rdm2_ehm ] again , the second dominating state is @xmath174 which overwhelms all other eigenstates except @xmath166 . from eq .
( [ eq : evec_ehm ] ) , we also notice that the state @xmath174 only consists of the charge part . as a result
, we may argue that in the this region , the charge part is decoupled from the spin part .
the reduced density matrix can be reduced to an effective one in eq .
( [ eq : rho_charge ] ) .
similarly , for @xmath175 , the spin part in @xmath166 outweigh the charge part as one can realize from fig .
[ fig : rdm_p_ehm ] . the eigenvalues for @xmath176 , @xmath177 , and @xmath178 also dominate in the low - lying eigenstates .
the spin degree of freedom is singled out in this regime and the reduced density matrix can be effectively expressed as the one in eq .
( [ eq : singlerdm ] ) . and @xmath3 in the 1d extended hubbard model .
@xmath42 , @xmath158 and @xmath25 .
( bottom ) a plot of the mutual information as a function of @xmath3 for @xmath179 .
the inset shows a ln - ln plot of the mutual information as a function of @xmath3 for @xmath180.,title="fig:",width=302 ] + and @xmath3 in the 1d extended hubbard model .
@xmath42 , @xmath158 and @xmath25 .
( bottom ) a plot of the mutual information as a function of @xmath3 for @xmath179 .
the inset shows a ln - ln plot of the mutual information as a function of @xmath3 for @xmath180.,title="fig:",width=302 ] in the 1d extended hubbard model . here
@xmath42 , @xmath158 and @xmath25.,width=302 ] in fig .
[ fig:3d_entro_ehm_u4 ] , the mutual information between two single sites are calculated as a function of @xmath14 and the separation @xmath3 . in the 2d plot
, we can see that the mutual information for @xmath181 is non - vanishing at a long distance . for @xmath182 ,
the ln - ln plot in the inset also shows that the mutual information decays algebraically with @xmath3 .
so the long - range correlation in the two regimes is already captured by a block size of one site . to construct the order parameter in these regimes
, we may go back to consider the single - site reduced density matrix .
figure [ fig : evrdm_ehm_u4 ] is a plot of the eigenvalues of the single - site reduced density matrix .
it is qualitatively the same as the case of @xmath110 .
the @xmath183 and @xmath184 states are dominated for @xmath182 while the @xmath185 and @xmath186 states are dominated for @xmath181 . following similar analysis in the previous sections ,
we can find that the order operators are @xmath187 and @xmath188 , which characterize the sdw and cdw respectively , in the two regions .
however , the above is not the whole story .
the single - site reduced density matrix was not enough to capture the correlation in the system around @xmath172 .
returning to fig .
[ fig : rdm2_ehm ] , the eigenvalue of @xmath166 has a drop around @xmath172 and there is also a relatively large rise in the weight of the eigenstates @xmath189 , @xmath190 , @xmath191 and @xmath192 .
in addition , the magnitude of @xmath171 and @xmath165 in @xmath166 becomes comparable in this intermediate value .
these observation suggest that the spin part and the charge part are coupled around @xmath172 .
there may exist some other kind of long - range correlation rather than sdw and cdw in this intermediate region .
let s consider @xmath159 again and rearrange the single - block reduced density matrix from the basis of @xmath193 to @xmath194 .
the purpose of doing this is that we want to filter out the spin - down degree of freedom from the spin - up degree of freedom ( or vice versa ) .
after the rearrangement , we traced out the spin - down degree of freedom .
the resulting reduced density matrix has the form of @xmath195 in the basis of @xmath196 .
similarly , if we trace out the degree of freedom for spin - up , we have @xmath197 in the basis of @xmath198 .
next , we would like to compare the off - diagonal matrix elements @xmath199 in @xmath200 and @xmath201 with that corresponding to sdw and cdw .
the weight of sdw is given by the coefficient of @xmath202 or @xmath203 ( the notation here is in form of @xmath204 ) . from eq .
( [ eq : evec_ehm ] ) , it can be obtained by @xmath205 where the @xmath206 s are the eigenvalues of the corresponding eigenstates in eq .
( [ eq : evec_ehm ] ) . for the weight of cdw
, we obtained it by considering the matrix element corresponds to @xmath207 or @xmath208 from eq .
( [ eq : evec_ehm ] ) .
we have @xmath209 .
@xmath42 , @xmath158 and @xmath159.,width=302 ] figure [ fig : rdm_order_ehm ] shows a plot of @xmath210 , @xmath211 and @xmath199 as a function of @xmath14 .
on the two sides far away from @xmath172 , the largest weight would correspond to sdw and cdw respectively as expected from previous analysis .
however , around @xmath172 , the off - diagonal matrix element in @xmath200 and @xmath201 are dominating instead .
if we just pick this dominating weight to define the order parameter , we have @xmath212 as mentioned before , the system possess up - down spin symmetry .
we could take @xmath213 , and then separate the operator into the real and imaginary part .
we have @xmath214 either the first term or the second term in the bracket above , or their linear combination can be taken as the order parameter .
let s take the real part as the order parameter , i.e @xmath215 in terms of the fermion creation and annihilation operators , we have @xmath216 which is the conventional order parameter that has been used to investigate the bow in the extended hubbard model @xcite .
we summarize our scheme of constructing the order parameter in the following : ( i ) calculate the mutual information to see whether there exists any long - range correlation in the system .
choose the minimum size of the block where the mutual information is non - vanishing for two blocks separated by a large distance .
( ii ) calculate the difference between the diagonal matrix elements of the two block reduced density matrix and the product of the diagonal matrix elements of two single block reduced density matrix .
this is to single out the correlating elements from other noise due to the finite size effect .
( iii ) obtain the eigenvalues and eigenvectors of the single block reduced density matrix . construct the order parameter from the heavy weighted eigenstates
apply traceless condition and cut - off condition to narrow down the choice .
( iv ) calculate the correlation function using the derived order parameter and study the mode of it .
with the scheme proposed , we have derived the order parameters which show long - range correlation in the ground state of the 1d ehm without using any empirical knowledge .
such an application confirmed that the order parameter for a quantum many - body system can be systematically derived even without the knowledge of symmetry in the system .
we expect that our scheme can shed new lights on the controversies in some frustrated antiferromagnet@xcite and bond - order wave in the extended hubbard model@xcite . throughout the paper ,
the system size being simulated was 14 sites .
remarkably , we see that even for such a small system size , the spectra of the single - site or two - site reduced density matrices were able to capture the information of the nature of the order in the ground state of the system .
this may be a promising scheme as one could have insight on the order existing in the system without simulating a large system , which requires intensive computational powers .
however , we would like to mention that the crossing of the dominating states in the reduced density matrix spectrum may not locate the phase boundaries exactly . to locate the phase boundaries
, one can use the order operator derived to calculate the correlation function and then perform finite size scaling analysis .
h. q. lin , e. r. gagliano , d. k. campbell , e. h. fradkin , and j. e. gubernatis , in _ the hubbard model : its physics and mathematical physics _ , edited by d. baeriswyl _ et .
315 - 327 ( plenum , new york , 1995 ) ; see also , h. q. lin , d. k. campbell , and r. t. clay , chin . j. phys . * 38 * , 1 ( 2000 ) . | without any knowledge of the symmetry existing in the system , we derive the exact forms of the order parameters which show long - range correlation in the ground state of the one - dimensional extended hubbard model using a quantum information approach .
our work demonstrates that the quantum information approach can help us to find the explicit form of the order parameter , which can not be derived systematically via traditional methods in the condensed matter theory . |
the gas scale height of galactic disks is an important quantity related to the process of star formation .
it can be used to constrain the surface density of matter , the balance between star formation processes and gravity , and the ism pressure , which must be important for regulating star formation .
optical images of edge on galaxies show that the scale height of stellar disks are independent of position along their major axis ( van der kruit & searle 1981 , 1982 ; kylafis & bahcall 1987 ; shaw and gilmore 1990 ; barnaby & thronson 1992 ) , or slowly increasing with radius ( de grijs & peletier 1997 ) .
the constancy of the stellar scale height must be related to the radial and vertical distribution of the gas , since molecular clouds ( mcs ) are responsible for both the formation of stars and the dynamical heating of the stellar disk ( spitzer & schwarzschild 1951 ) .
the vertical distribution of neutral hydrogen in our galaxy has been studied by many authors .
it is found that the hi scale height @xmath1 is approximately constant over a large range of galactocentric radii ( dickey & lockman 1990 ; heiles 1991 ) .
it is commonly assumed that the hi gas is supported in the galactic gravitational potential primarily by turbulence ( lockman & gehman 1991 ) , with contributions from magnetic and cosmic ray pressure ( parker 1966 ; spitzer 1990 ) and radiation ( ferrara 1993 ) .
the constancy of the gas scale height in galactic disks is a common assumption in several models .
however , it should not be taken for granted , since direct measurements of the hi scale height in external galaxies are difficult and sparse , and the hi scale height has never been mapped in a face on galaxy . a nearby edge - on galaxy where the hi distribution has been studied with very high linear resolution is ngc 891 .
sancisi & allen ( 1979 ) found the hi disk of ngc 891 to be less than 1 kpc thick near the center , increasing outwards up to 12 kpc .
the hi disk flare is not confirmed in the later study by swaters , sancisi & van der hulst ( 1997 ) , where a significant hi halo component is instead identified , and it is pointed out that the inferred hi vertical density profile is dependent on the adopted kinematic model . the co scale height for the same galaxy ( @xmath2 pc ) has also been estimated ( scoville et al .
1993 ) . in this letter we propose a new method to measure and map the hi scale height of face on galaxies .
the method is based on the application of the spectral correlation function ( scf ) to hi interferometric surveys .
the scf ( rosolowsky et al . 1999 ; padoan , rosolowsky & goodman 2001 ) quantifies the correlation between spectra at different map positions as a function of their separation , and is sensitive to the properties of both the gas mass distribution and the gas velocity field .
the scale corresponding to the scale height of the hi disk should mark the transition from small scale three dimensional turbulence to two dimensional large scale motions on the disk , since the ratio between radius and scale height of the cold gas in typical disk galaxies is very large ( @xmath3 ) .
this transition should appear as a feature in the scf at a spatial lag approximately equal to the hi scale height .
on smaller scales the scf should instead be a power law reflecting the self similar nature of turbulence , as in mcs ( padoan , rosolowsky & goodman 2001 ) .
we have applied this method to the australia telescope compact array ( atca ) interferometric hi survey of the large magellanic cloud ( lmc ) by kim et al .
( 1998a , b , 1999 ) , combined with the single dish survey done with the parkes multibeam by staveley
smith et al .
( in preparation ) .
results of the combined survey are presented in kim et al .
( in preparation see also astro
ph/0009299 ) .
we find that the scf is indeed a power law on small scale , and has a characteristic feature at approximately 180 pc , where the power law is interrupted .
we identify the scale of the clear `` break '' in the scf as the hi scale height .
we are also able to map significant variations of the scale height over the disk , between @xmath4 and 240 pc . in a recent work , elmegreen , kim & staveley smith ( 2001 )
have computed the power spectrum of hi intensity maps of the lmc from the same survey used in this work .
they find that the power spectrum of intensity is steeper on small scales than on large ones .
the transition occurs at approximately 80100 pc , and is interpreted as the line of sight thickness of the cold hi layer .
the present letter is qualitatively a confirmation of this earlier result , in the sense that both works show the possibility of using the statistical information from hi interferometric surveys to estimate the hi scale height of a face on galaxy .
the spectral correlation function ( scf ) measures the correlation between spectra at different map positions as a function of their spatial separation , and is sensitive to the properties of the gas mass distribution and the gas velocity field ( rosolowsky et al .
1999 ; padoan , rosolowsky & goodman 2001 ) .
let @xmath5 be the antenna temperature as a function of velocity channel @xmath6 at map position @xmath7 .
the scf for spectra with spatial separation @xmath8 is : @xmath9 where the average is done over all map positions @xmath7 .
@xmath10 is the scf uncorrected for the effects of noise , @xmath11 ^ 2 } { \sigma_vt(\vecr , v)^2+\sigma_vt(\vecr+\vecdr , v)^2 } } \right\rangle _ { \vecdr } , \label{2}\ ] ] where the average is limited to separation vectors @xmath12 with @xmath13 , and @xmath14 is the scf due entirely to noise , @xmath15 and @xmath16 is the `` spectrum quality '' ( see discussion in padoan , rosolowsky & goodman 2001 ) .
@xmath16 is computed as the ratio of the rms signal within a velocity window @xmath17 to the rms noise , @xmath18 , @xmath19 where @xmath20 is the width of the velocity channels .
the scf has been computed for both molecular cloud data and synthetic data obtained by solving the radiative transfer through the three dimensional density and velocity fields of numerical simulations of super sonic mhd turbulence .
the result is typically a power law that extends up to a separation @xmath8 comparable to the map size , reflecting the self
similarity of super sonic turbulence ( see padoan , rosolowsky & goodman 2001 , and references therein )
. = 1 = 8.2 cm = 1 = 8.7 cm = 1 = 8.2 cm
we have computed the scf for several regions of the australia telescope compact array ( atca ) interferometric hi survey of the large magellanic cloud ( lmc ) by kim et al .
( 1998a , b , 1999 ) , combined with the single dish survey done with the parkes multibeam by staveley
smith et al .
( in preparation ) .
the original data has been rebinned to a resolution of 80 , that is a linear resolution of 20 pc assuming a distance to the lmc of 50 kpc ( feast 1991 ) .
we therefore obtain a map made of @xmath21 spectra ( @xmath22 kpc ) , with 120 velocity channels .
the velocity resolution is 1.65 km / s .
details about the atca survey can be found in kim et al .
( 1998a , b , 1999 ) .
the finite size of a map with non periodic boundaries generates edge effects in correlation functions . in order to avoid this problem ,
we have computed the scf for subsets of @xmath23 spectra ( @xmath24 kpc ) , excluding the data from the `` edges '' of the map .
the spectra in each subset have been compared with all spectra within a larger area around that subset , including @xmath25 spectra ( @xmath26 kpc ) . in this way
the scf for all the spectra in each subset can be computed up to a separation equal to the size of the subset ( 1 kpc ) , without any effects related to the finite size of the map .
figure 1 shows the integrated hi intensity map of the lmc , divided into @xmath22 subsets with @xmath23 spectra each .
the scf has been computed only for the @xmath27 central ones ( delimited by the darker square ) for which complete sets of @xmath25 reference spectra are available .
for lack of space , we have plotted in figure 2 only the scf computed inside the area marked with the rectangle in figure 1 .
figure 2 shows that the scf is a power law on small scales over almost an order of magnitude range in spectral separation @xmath8 .
the power law slope of the scf varies from -0.40 to -0.15 over the whole map .
the power law is very clearly interrupted at @xmath0 pc , where the slope of the scf generally decreases . in 22 of the 25 regions ,
the scf breaks to a `` shallower '' slope on larger scales .
it sometimes becomes steeper again on the scale of a few hundred pc , due probably to the effect of differential rotation , which grows with increasing spectral separation . in only 3 of the 25 cases ,
the scf breaks from a power law to a `` steeper '' slope .
this happens in the 3 regions around the location of 30 doradus , and south of it ( the 3 regions covering the largest intensity feature on the bottom left of the map in figure 1 ) .
we interpret the interruption of the power law shape of the scf as the outer scale of three dimensional super sonic turbulence in the lmc disk .
a scale free cascade of three dimensional turbulence is unlikely to extend to scales much larger than the disk scale height , as virtually two dimensional flows should dominate the dynamics on those larger scales .
the power law shape of the scf for the hi data reflects the presence of self similar turbulence , probably extending to much smaller scale than probed by the present survey ( 20 pc in our rebinned data ) .
below 20 pc the turbulent cascade is well probed by the scf of molecular emission line maps ( see padoan , rosolowsky & goodman 2001 )
. assuming that the the scf is indeed sensitive to the hi scale height , @xmath1 , we find that @xmath1 varies in the range @xmath28 pc over the surveyed area .
the largest value is found in the region around 30 doradus .
statistical error bars are smaller than the size of the symbols used in the plots in figure 2 , and variations of @xmath1 are therefore significant .
figure 3 shows a contour map of @xmath1 over the @xmath27 regions where the scf has been computed .
the average value is @xmath29 pc .
kim et al . ( 1999 ) estimated the hi scale height using the average vertical velocity dispersion and the average surface density of both the hi and the stellar components of the lmc disk , and adopting the disk model by dopita & ryder ( 1994 ) .
they found @xmath30 pc .
this is almost identical to the value of @xmath31 estimated with the scf method , which is the main reason why we interpret the outer correlation scale of the scf as the local value of the disk scale height .
nevertheless , we can not rule out possible alternative interpretations of the scf results , and the subject deserves further study .
elmegreen , kim & staveley smith ( 2001 ) have recently computed the power spectrum of hi intensity maps of the lmc obtained with data from the same survey used in this work .
they find a steeper power spectrum on small scales than on large ones .
the transition occurs at approximately 80100 pc , and is interpreted as the line of sight thickness of the hi layer , based the theoretical results by lazarian & pogosian ( 2000 ) . in this work
we confirm the possibility of using the statistical information from hi interferometric surveys to estimate the hi scale height of a face on galaxy .
however , there is a fundamental difference between the present letter and the work by elmegreen , kim & staveley smith ( 2001 ) , apart from the different method of data analysis adopted .
elmegreen , kim & staveley smith ( 2001 ) implicitly assume that the three dimensional power spectrum of the turbulent density field is described by a unique power law over all scales , and the change in slope of the two dimensional power spectrum of integrated intensity is merely a projection effect . here
instead we argue that there must be a physical transition in the statistical properties of the flow on a scale close to the disk scale height , which the scf is sensitive to .
we have proposed a new way to measure and map the scale height of the gas in face on disk galaxies , based on the application of the scf to hi interferometric surveys .
we have shown that this method applies successfully to the lmc .
it can be applied to a number of nearby galaxies , depending on the spatial and velocity resolution of hi or molecular transition interferometric surveys .
the bima survey of nearby galaxies , bima song ( regan et al . 2001 ) , for example , has an angular resolution of approximately 6 ( using the d and c array configurations ) , and can achieve approximately 3 , with the b array configuration . at the distance of 4 mpc , a 3 resolution corresponds to a linear resolution of @xmath32 pc .
there are six galaxies within the approximate distance range between 2 and 4.5 mpc , in the bima song list : ic0342 , ngc2403 , ngc2976 , ngc3031 , ngc4736 and ngc4826 .
they could in principle be studied with this method .
problems with the co surveys may arise due to the low surface filling factor of co relative to hi , and , in the particular case of the bima song survey , due to the relatively low velocity resolution ( 4.06 km / s ) .
however , a velocity resolution of approximately 1.0 km / s could be easily achieved with bima in these galaxies ( tamara t. helfer , private communication ) . for 21-cm hi observations , a 50-km baseline in a radio interferometer
gives a resolution of 16 pc at 4 mpc . at 3-mm
, co gives the same kind of resolution on a 0.7-km baseline .
therefore , the synthesized beams of the immediate descendants of existing cm and mm interferometers ( e.g. the evla and carma ) should be able to map many galaxies with more than adequate resolution for the scf to be used to measure , and map out , galactic scale height .
the application of this method to a number of galaxies could provide new important measurements of the gas scale height and its variations inside individual disks and between different galaxies .
once the scale height is measured , it can also be used , together with the observed vertical velocity dispersion , to put new constraints on the distribution of dynamical mass in galactic disks . * figure captions : * + * figure [ fig1 ] : * velocity integrated hi intensity in the lmc from the atca interferometric survey by kim et al .
( 1998a , b , 1999 ) , combined with the single dish survey done with the parkes multibeam by staveley
smith et al .
( in preparation ) .
the perimeter of each of the 25 regions used to compute the scf is over over plotted . + * figure [ fig2 ] : * the scf for the third and fourth column of regions from figure 1 .
the regions are given coordinates ( x , y ) starting from ( 0,0 ) at the bottom left corner of the @xmath27 square in figure 1 . + * figure [ fig3 ] : * contour map of the scf scale height of the lmc disk .
values of contour lines are in pc .
the intensity map shows the hi integrated intensity as in figure 1 . | we propose a new method to measure and map the gas scale height of nearby disk galaxies .
this method is applied successfully to the australia telescope compact array interferometric hi survey of the large magellanic cloud ( lmc ) ; it could also be applied to a significant number of nearby disk galaxies , thanks to the next generation of interferometric facilities , such as the extended vla and carma .
the method consists of computing the spectral correlation function ( scf ) for a spectral line map of a face on galaxy .
the scf quantifies the correlation between spectra at different map positions as a function of their separation , and is sensitive to the properties of both the gas mass distribution and the gas velocity field .
it is likely that spatial correlation properties of the gas density and velocity fields in a galactic disk are sensitive to the value of the scale height of the gas disk .
a scale free turbulent cascade is unlikely to extend to scales much larger than the disk scale height , as the disk dynamics on those larger scales should be dominated by two dimensional motions .
we find a clear feature in the scf of the lmc hi disk , on the scale of @xmath0 pc , which we identify as the disk scale height .
we are also tentatively able to map variations of the scale height over the disk .
epsf 2ns^/n |
the four nucleon bound state calculations have , in the last years , reached a high level of accuracy and consistency at least as far as the solutions of the corresponding equations are concerned @xcite .
this situation contrasts with the 4n scattering problem where despite some pioneering and relevant results @xcite , there is a manifest lack of convergence among the different groups and methods even when using simple interactions .
this problem is not only a general extension of the 3-body one , in the sustained task of the nuclear few - body community to deal with increasingly complex systems , but we believe it constitutes a qualitative jump in our understanding of nuclear systems . indeed
the continuum spectrum of the 4n system ( see figure [ figchart ] ) , with its rich variety of thresholds and structures , provides a bridge between the relative simplicity of the a=2,3 problems and the complexity of many - body systems .
even when restricted to the energies below the first 3-body breakup threshold , the presence of several resonances at each z channel , the existence of the almost degenerate p+t and n+@xmath1he thresholds with , in the middle , the first @xmath2 excitation of the @xmath3he ground state make the understanding of the a=4 chart in terms of fundamental nn interactions an exciting and redoubtable theoretical challenge .
we present here the first solution of the faddeev - yakubovsky ( fy ) equations in configuration space for the four nucleon scattering problem . although the results concerning the bound states ( @xmath3he and @xmath3he@xmath4 ) will be discussed in some detail ,
our main interest lies in the 4n continuum spectrum , i.e. the n+nnn elastic scattering and its coupling to the first inelastic nn+nn threshold .
the resolution method is based on the angular momentum expansion of the fy amplitudes and the spline expansion of their radial parts .
orthogonal collocation is used to generate a linear system which is solved by iterative procedures .
the scattering observables are extracted from a direct inspection of the fy amplitudes in the asymptotic region , in a natural extension of the methods developed for the three - body case in @xcite .
the results presented in this paper have all been obtained by using an s - wave nn interaction model and the isospin symmetry hypothesis .
the coulomb and mass difference effects are thus not included .
this choice , guided by methodological reasons , allows a presentation of the formalism and methods in a relatively simple framework .
is is remarkable , however , that such a simple model provides a very good description of low energy scattering observables even if , as in the n+t case , they are not totally trivial .
some first results including realistic interactions have already been reported elsewhere @xcite and will be the subject of subsequent publications .
the paper is organized as follows . in the next section
we describe the general formalism and the simplifications arising in the case of 4 identical particles .
this section contains also the spin , isospin and angular momentum algebra . in section 3
we give some details of the numerical methods used . in section 4
the results will be presented .
they include the @xmath3he ground and first excited state , the elastic phase - shifts and low energy parameters for the n+nnn reaction and the n+nnn @xmath5 nn+nn first inelastic channel .
the energies are restricted below the 3- and consequently 4-body break - up .
conclusions and perspectives will be given in the last section .
with the aim of solving the schrdinger equation for n particles interacting via a pairwise potential @xmath6 @xmath7 yakubovsky @xcite , generalizing faddeev s work for n=3 @xcite , wrote a set of equations whose solutions verify ( [ sch ] ) and which provides a proper mathematical scheme to account for the variety of physical situations involved ( see , e.g. figure [ voies ] ) . in the n=4 case ,
the fy equations can be obtained by first splitting the total wavefunction @xmath8 in the usual faddeev amplitudes , @xmath9 , associated with each interacting pair : @xmath10 and requiring them to be a solution of the system of coupled equations @xmath11 or equivalently in its integral form @xmath12 with @xmath13 each amplitude @xmath9 is in its turn split in 3 parts , the fy amplitudes , corresponding to the different asymptotics of the remaining two particles : @xmath14 and obeying the following system of coupled equations : @xmath15 in which an amplitude @xmath16 , not defined by ( [ deffy ] ) , has to be understood as being identical to @xmath17 .
any solution of this system of 18 coupled equations , called the fy equations , is a solution of ( [ fed ] ) and consequently of the initial problem ( [ sch ] ) .
its advantage lies in the possibility to define for system ( [ fyde ] ) appropriate boundary conditions ensuring the unicity of the solution .
indeed when one of the particle , e.g. labeled by @xmath18 , is out of reach of the interaction , all the amplitudes in ( [ fyde ] ) tend to zero except @xmath19 and circular permutations on @xmath20 which obey @xmath21 this system of equations , resulting from ( [ fyde ] ) , is equivalent to the 3n faddeev equations for the particles @xmath22 . in a similar way , when the @xmath23 and @xmath24 clusters are free from interaction the only non vanishing amplitudes are @xmath25 and @xmath26 and their corresponding equations in ( [ fyde ] ) tends to @xmath27 it is worth noticing that the fy amplitudes can be written in terms of the faddeev amplitudes in the form : @xmath28 where @xmath29 and , according to ( [ fei ] ) , in terms of the total wavefunction : @xmath30 disregarding the internal degrees of freedom ( like spin , isospin ) , the natural basis for the configuration space is provided by the positions of the different particles @xmath31 in order to remove the center of mass motion , it is useful to introduce the relative jacobi coordinates .
two sets of jacobi coordinates can be defined for each of the 4 !
arrangements @xmath32 .
one of k type : @xmath33 and one of h type : @xmath34 in which @xmath35 is an arbitrary mass taken as a reference , and @xmath36 is the reduced mass of clusters @xmath37 and @xmath38
. however some of these 48 coordinate sets are redundant .
for instance those obtained by exchanging @xmath39 in a k - set or @xmath39 or / and @xmath40 in the h - set are equivalent .
this yields 18 ( 12k+6h ) arbitrary and physically non equivalent jacobi sets , as many as fy amplitudes .
any of these coordinate sets , suitably completed with the center of mass coordinate @xmath41 , constitutes an equivalent description of the 4 particles configuration space .
that provides 18 coordinate sets and the corresponding bases for the configuration space , equivalent to ( [ baser ] ) , that will be written in the form @xmath42 or @xmath43 .
the degrees of freedom related to the center of mass motion separate in nonrelativistic dynamics and will be hereafter omitted .
although each fy amplitude could be in principle expressed in terms of any of these bases , only one of them is appropriate for expanding it .
we will denote the resulting components by : @xmath44 the bases described above have to be completed to account for other degrees of freedom like spin , isospin , etc .
further details about the formalism and the relation between the different bases sets can be found , e.g. , in @xcite . in the case of four identical particles ,
the 18 fy amplitudes can be obtained by the action of the transposition permutation operators @xmath45 on two of them , arbitrarily chosen provided that one is of k type and the other one of h type .
we have taken @xmath46 and @xmath47 .
the 4-body problem is solved by determining the two @xmath48 amplitudes which satisfy the following equations : @xmath49\label{fye1}\\ ( e - h_0-v ) h&=&v\left [ ( p_{13}p_{24}+p_{14}p_{23 } ) \ ; k + p_{13}p_{24 } \ ; h\right ] \label{fye2}\end{aligned}\ ] ] where @xmath50 depending on whether the particles are bosons or fermions .
the asymptotic equations , i.e. the equivalent of ( [ fya1],[fya2 ] ) are in this case @xmath51 the total wave function is then given by : @xmath52\ ; \left [ 1 + \varepsilon(p_{14}+p_{24}+p_{34 } ) \right]k \label{psi13}\\ \psi_{2 + 2 } & = & \left [ 1+\varepsilon ( p_{13}+ p_{23}+ p_{14}+ p_{24 } ) + p_{13}p_{24}\right ] \;h \label{psi22}\end{aligned}\ ] ] each amplitude @xmath53 is considered as a function of its natural set of jacobi coordinates @xmath54 , defined respectively by equations ( [ k ] ) and ( [ h ] ) with @xmath55 and @xmath56 : @xmath57@xmath58 they are expanded in angular momentum variables for each coordinate according to @xmath59 where @xmath60 are generalized tripolar harmonics containing spin , isospin and angular momentum variables and the functions @xmath61 , the reduced radial fy components , are the unknowns .
the label @xmath37 represents the set of intermediate quantum numbers defined in a given coupling scheme and includes the specification for the type of amplitudes ( k or h ) .
we have used the following couplings , represented in figure [ coupling ] : @xmath62_{t_3 } t_4 \right\}_t \otimes \left\ { \left [ \left ( l_x ( s_1 s_2)_{\sigma_x } \right)_{j_x } ( l_y s_3)_{j_y } \right]_{j_3 } ( l_z s_4)_{j_z } \right\}_{j^{\pi } } \cr \mbox{h amplitudes } & : & \left [ ( t_1 t_2)_{\tau_x } ( t_3 t_4)_{\tau_y } \right]_t \otimes \left\ { \left [ \left ( l_x ( s_1 s_2)_{\sigma_x } \right)_{j_x } \left ( l_y ( s_3 s_4)_{\sigma_y } \right)_{j_y } \right]_{j_{xy } } l_z \right\}_{j^{\pi } } \end{aligned}\ ] ] where @xmath63 and @xmath64 are the spin and isospin of the individual particles and @xmath65 are respectively the total angular momentum , parity , and isospin of the 4-body system .
each component @xmath61 is thus labeled by a set of 12 quantum numbers to which the symmetry properties of the wavefunction impose the additional constraints : @xmath66 for k and @xmath67 for h. the total parity @xmath68 is given by @xmath69 in both coupling schemes .
the radial equations for the components @xmath61 are obtained by projecting each of equations ( [ fye1]-[fye2 ] ) in its natural configuration space basis @xmath70 .
several steps further @xcite we end with a system of coupled integro - differential equations which , most generally , can be written in the form : @xmath71 \label{fse}\end{aligned}\ ] ] with @xmath72 the functions @xmath73 contain all the spin , isospin and angular momentum couplings .
the arguments @xmath74 are functions of @xmath75 in the more general case , and are detailed in the appendix .
the system of equations ( [ fse ] ) has been explicitly written in @xcite for the case of four identical bosons .
the fy components for the different ( s , t ) channels in the s - wave approximation , i.e. with all orbital angular momenta in expansion ( [ kpw ] ) equal to 0 , are listed in table [ tab_amplis ] . in this table ,
the symbols `` @xmath5 '' and `` @xmath76 '' denote respectively the amplitudes corresponding to an asymptotic n+nnn or nn+nn channel .
note that , contrary to the 3n problem , the number of fy components appearing in ( [ fse ] ) is infinite even when the pair interaction is restricted to a finite number of partial waves .
this divergence comes only from the existence of the @xmath77 additional degree of freedom in the k - like amplitudes .
for all the physical problems we have considered , the boundary conditions can be written in the dirichlet form .
the use of reduced radial fy components @xmath61 in ( [ kpw ] ) imposes for any kind of solution the regularity conditions : @xmath78 for the bound state problem these conditions are completed by forcing the components @xmath61 to vanish on the hypercube @xmath79\times[0,y_n]\times[0,z_n]$ ] , i.e : @xmath80 for the scattering problems the boundary conditions are implemented by imposing at large enough values of @xmath81 the asymptotic behavior of the solutions .
thus , for the n+nnn elastic case we impose at @xmath82 the solution of the 3n problem for all the quantum numbers @xmath83 corresponding to the open asymptotic channel @xmath84 @xmath85 where the functions @xmath86 are the faddeev amplitudes of the 3n problem .
indeed , at large values of @xmath81 and for energies below the first inelastic threshold , the solution of equation ( [ fyai1 ] ) factorizes into a bound state solution of the 3n faddeev equations and a plane wave in the @xmath81 direction with momentum @xmath87 , whereas the solution of ( [ fyai2 ] ) vanishes .
one then has , e.g. for an s - wave , @xmath88 and the scattering observables are directly extracted from the logarithmic derivative of the k amplitude in the asymptotic region : @xmath89 provided we are in the asymptotic domain , the phase shifts thus extracted have to be independent of @xmath90 and of the amplitude index @xmath83 .
this provides a strong numerical test .
an additional advantage of this procedure is that it avoids any cumbersome multidimensional integrals . in the presence of several open channels , like n+nnn and nn+nn e.g. ,
several resolutions are needed .
the boundary conditions ( [ asym1 ] ) are simply generalized in the form @xmath91 in which @xmath92 are arbitrary real numbers and the functions @xmath93 coincide with @xmath94 if @xmath84 is a n+nnn channel or are analogous to the faddeev amplitudes for ( [ fyai2 ] ) if @xmath84 is a nn+nn channel .
equation ( [ asym1 ] ) is generalized in the following way : @xmath95 where @xmath96 are the regularized hankel functions @xcite , and @xmath97 is a multiplicity number for the channel @xmath84 ( @xmath98 for 1 + 3 and @xmath99 for 2 + 2 channels ) due to the symmetry properties of the total wave function ( see , e.g. @xcite ) .
the coefficients @xmath100 ( amplitudes of the incoming waves ) and the s - matrix elements are the unknowns .
they are obtained by identifying with the asymptotic form ( [ smatrix ] ) the values @xmath101 at @xmath82 for different solutions corresponding to different choices of @xmath102 s whose number equals the number of channels .
we remark that the momenta @xmath87 appearing in equations ( [ dlogphi],[qcotd],[smatrix ] ) are the conjugate variables of the @xmath81-jacobi coordinates .
they are related to the center of mass kinetic energy @xmath103 of channel @xmath84 according to @xmath104 .
the physical momenta , conjugate to the physical intercluster distances , are @xmath105 or @xmath106 depending whether @xmath84 corresponds to a 1 + 3 or 2 + 2 channel . by the definition ( [ smatrix ] )
one has unitarity ( @xmath107 ) and symmetry ( @xmath108 ) relations . working with real solutions these properties are related ( unitarity implies symmetry ) .
however none of them is a trivial consequence of the method used but a strong test of numerical accuracy . in order
that the factorizability takes place the asymptotic 3n or 2n+2n states have to be calculated with the same numerical scheme as that used to solve the 4-body problem .
this means in practice that they are exact solutions of ( [ fyai1 ] ) or ( [ fyai2 ] ) , once the z - dynamics are removed . by doing so
, the factorization property , valid only in cartesian coordinates , is an exact numerical property and leads to stable local results .
this behavior is illustrated in figure [ represnt ] in which the fy amplitudes for an n+nnn elastic scattering at @xmath109 and @xmath110 are represented .
the numerical methods used are based on the hermite spline expansion , orthogonal collocation @xcite and iterative procedures for solving the linear system .
an important step in their solution is the tensor trick @xcite .
we look for the solutions @xmath61 of the integro - differential system ( [ fse ] ) in the form of a tensor product of 1-dimensional cubic hermite splines @xmath111 : @xmath112 defined on grids of non - equidistant @xmath113 points g@xmath114 where @xmath115 .
a grid g@xmath116 will be defined by giving the number of intervals @xmath117 , the end point @xmath118 and the constant scaling factor between two consecutive intervals , @xmath119 .
we will use the following notation g@xmath120 , often extended to multi - domain grids according to g@xmath121 the boundary conditions are easily implemented using the properties of the spline functions ( value and derivative equal to 0 or 1 at the grid points ) .
they result in fixing the values of some of the unknown coefficients @xmath122 in the expansion ( [ sol1 ] ) . by exchanging the indices of the two last spline functions in each variable
the solution takes the form : @xmath123 where @xmath124 are the coefficients of the asymptotic functions @xmath125 defined in ( [ asym2 ] ) . in the cases of a closed channel or a bound state these coefficients
are zero .
collocation points associated with the 3-dimensional grid are used to generate a linear system , the @xmath126 being the unknowns .
the integral terms in ( [ fse ] ) are calculated with a gaussian quadrature rule with typically @xmath127 points in each angular variable @xmath128 and @xmath129 . in order to limit the number of parameters we have chosen @xmath130 .
the number of unknowns is given by @xmath131 where @xmath132 is the number of fy components .
a rough estimation for the extreme cases of a four bosons bound state and of a scattering state in a realistic problem leads to values @xmath133 , @xmath134 , @xmath135 and consequently to a number of unknowns going from @xmath136 to @xmath137 .
this implies the use of a dense matrix with @xmath138 coefficients .
direct methods are not appropriate for solving such huge linear systems , and we have used iterative techniques , which avoid any storage or inversion of the matrix .
the basic feature of any iterative method is to obtain the solution of the linear system only by iterative application of the matrix to an initial vector .
this implies a complete calculation of the matrix elements each time it is necessary , and requires the intensive use of parallel computers .
in the case of scattering states , the boundary conditions ( [ asym2 ] ) , responsible for the second term in ( [ phispline ] ) , generate a source term in ( [ fse ] ) leading to a regular linear system of the type @xmath139\vec{c}+\vec{b}$ ] , where the different @xmath140 matrices are reminiscent of the @xmath141 operators . for brevity we will write the system in the form @xmath142 .
the numerical method we have chosen to solve this system is gmres ( for generalized minimum residual algorithm ) @xcite .
gmres is a prototype of the so called krilov subspaces projection methods .
its aim is to minimize the residue @xmath143 of an approximate solution @xmath144 , starting from a trial vector @xmath145 and looking for its correction @xmath146 in the krilov subspace @xmath147 such that the residue @xmath148 is orthogonal to @xmath149 .
when the dimension @xmath150 increases , the residue of the approximate solution @xmath144 can be brought to an arbitrary small value , called tolerance . in most of the practical cases , a tolerance between @xmath151 and @xmath152 is sufficient . gmres is a powerful tool when the problem is well conditioned , what is almost never true in a realistic case .
the way out is to `` precondition '' the system , i.e. to solve the equivalent problem @xmath153 instead of @xmath142 .
the closer to @xmath154 is the matrix @xmath155 , the better is the preconditioning .
our choice was to take the matrix @xmath155 equal to @xmath156 .
as pointed out in @xcite its tensor structure , optimized by the choice of cartesian coordinates , allows an easy inversion .
our preconditioning technique gives us a converged result after @xmath157 matrix applications , for all the considered physical cases and independently of the dimension of the matrix .
examples of convergence curves , i.e. the evolution of the residue modulus at each step , are shown in figure [ evolres ] . in the case of bound state
, the asymptotic behavior of the wave function and fy amplitudes leads to a singular system @xmath139\vec{c}$ ] that will be rewritten for convenience in the form @xmath158 , the energy e being an eigenvalue .
it is well known that iterative methods are appropriate for extracting a few eigenvalues , but only those with largest modulus . with the 3-dimensional box - like boundary condition ( [ bsbc ] ) , the existence of overwhelming highly positive eigenvalues leaves no hope of obtaining directly the binding energy .
a crafty trick is to solve the eigenproblem @xmath159\vec{c}=\lambda\vec{c}$ ] where @xmath160 is now a parameter @xcite .
the value @xmath160 will be an eigenvalue of the initial problem when the spectrum @xmath161 contains 1 .
furthermore , it can be shown by variational considerations that the more excited the state ( including the non physical box states ) , the smaller the corresponding @xmath102 .
thus the eigenvalues of physical interest can be obtained with methods like ira ( implicit restarted arnoldi algorithm ) @xcite .
we notice that the full inversion of @xmath156 , including the two - body potential , gives a better conditioned spectrum @xmath161 than an inversion of its kinetic term alone , and avoids some of the awkward negative eigenvalues generated by the repulsive part of the potential .
ira is a generalization of the power method and gives the first eigenvalue in @xmath162 matrix applications .
it is also based on krilov subspaces , and approximates the eigenvalues of @xmath163 by those of the restriction of @xmath163 to the space spanned by @xmath164 vectors @xmath165 , @xmath166 being a trial arbitrary vector .
nevertheless , this method requires several calculations for different values of the energy and becomes numerically unstable when using highly repulsive two - body potentials , like the interatomic @xmath3he-@xmath3he one @xcite .
a more robust technique was finally adopted , often referred as shift - invert method .
the initial problem is written in the form @xmath167 and this technique converges to the energy closest to @xmath168 .
it gives very good results with a well - balanced mixture of ira and gmres .
ira is used to quickly obtain the dominant eigenvalue @xmath169 and provides the energy @xmath170 .
the real difficulty lies in the generation of the corresponding krilov subspace .
it is obviously impossible to invert @xmath171 since the @xmath172 matrices are contained in @xmath163 .
the step @xmath173 is then performed by a gmres resolution of the equivalent linear system @xmath174 .
this technique presents considerable interest especially for excited states which are easily obtained , independently of the previous convergence of the less excited ones .
the results presented in this section have been obtained with the spin - dependent s - wave interaction mt i - iii : @xmath175 the potential parameters and the value @xmath176 mev.@xmath177 are the same than those used e.g. in @xcite and are slightly different from those given in the original version @xcite . despite its bare simplicity , this potential turns out to be very efficient in describing the bulk of low energy properties in the few - nucleon systems .
we will first examine what we call the s - wave approximation , i.e. the fact that aside from the zero angular momentum of the interaction pair , all the angular momenta variables in expansion ( [ kpw ] ) are set equal to zero .
the convergence with respect the @xmath178 expansion will be examined in a second step . in the 4n - system ,
the bound states exist only for the s = t=0 channel . in the s - wave approximation
the number of fy components is limited to @xmath179 ( see table [ tab_amplis ] ) .
the binding energies and r.m.s .
radius for the ground ( @xmath3he ) and first excited ( @xmath3he@xmath4 ) states are given in table [ tab_bs ] .
the corresponding grids are g@xmath180 with @xmath181 and g@xmath182 with @xmath183 , given in table [ tab_grids1 ] .
the estimated accuracy in the binding energies is 0.01 mev but we notice that much less expensive calculations can provide a precise result as well ; e.g. the grid g@xmath184 with @xmath185 gives also a binding of 30.30 mev . in the ground state
we remark a good agreement with the best existing calculations @xcite .
the first excitation which , experimentally , corresponds to a @xmath186 resonance 0.40 mev above the p+t threshold @xcite , appears in this model as a loosely bound state .
the binding energy with respect to the n+nnn threshold ( e=-8.535 mev in this model ) is 0.257 mev .
a similar result was found in @xcite in which different versions of the one term separable yamaguchy potential gives a binding energy varying from 0.07 to 0.40 mev , depending on their different d - state contribution .
this @xmath2 first excitation has been widely considered in the literature as being a breathing - mode @xcite .
our conclusion is however different .
we have calculated the regularized two - body correlation functions defined by @xmath187 where @xmath188 represents the total wave function component in the @xmath189 quantum numbers , and where @xmath190 denotes the subset of quantum numbers @xmath37 relative to the @xmath191 variable ( @xmath192 ) .
the summation in ( [ correlfunct ] ) is performed onto one of the two basis , k or h. once the total wavefunction is normalized , the correlation functions are normalized according to : @xmath193 the results are displayed in figure [ figcorrel ] .
the separated contributions from the singlet and triplet state are plotted for ( a ) triton ( b ) @xmath3he ground state and ( c ) @xmath3he first excited state .
the difference between the correlation function for the ground and excited states is remarkable , both in the shape and in the separated singlet - triplet contributions . for the excited state one can distinguish the superposition of two structures with two different length scales , the short distance part being similar to the triton one
this suggests a 1 + 3 structure for the @xmath3he excited state , as can be more clearly seen in plot ( d ) where the results of ( c ) are compared with those of the triton suitably normalized .
contrary to what would happen in a breathing mode , the short distance behavior of the nucleons is that of an unperturbed triton with the fourth nucleon being simply a remote spectator . by modifying the mt i -
iii potential strength we failed to pull the state out of the bound region , the 3-body threshold moving in the same direction .
it seems very difficult for a pure strong interaction to generate a first excitation in the continuum .
the right position of this resonance is however a crucial point in any attempt to describe the low energy data ( e.g. p+t ) @xcite .
the effect of coulomb interactions could be enough .
however the inclusion of an ad - hoc repulsive 4-body term @xmath194 can also achieve the same result .
the preceding results are only slightly modified by the inclusion of higher partial waves in the fy expansion .
the effects of these contributions can be seen in table [ tab_bs2 ] .
their smallness shows the validity of the s - wave approximation .
these results have been obtained using the grid g@xmath184 with @xmath185 for the ground state and g@xmath182 with @xmath183 for the first excitation for which the corresponding triton binding energy is @xmath195 mev .
a crucial point in our method is to ensure the proper description of the 3n asymptotic state .
this is used to fix the grid parameters for the @xmath196 variables . in order to exhibit the stability of our results
, we will compare the phase - shifts obtained with several tritons corresponding to increasing numerical accuracies .
the considered grids , detailed in table [ grids2 ] , are t@xmath197 with @xmath183 and a binding energy @xmath198 mev , t@xmath199 with @xmath185 and @xmath200 mev , and t@xmath201 with @xmath185 and @xmath202 mev .
we recall that the precise value for the 3n binding energy is @xmath202 mev .
the grid parameters for the z - variable depends substantially on the relative kinetic energy .
a zero energy calculation requires a relatively large value of @xmath203 but very few points inside are sufficient to describe an asymptotic linear behavior . on the contrary
as far as the energy increases , the value of @xmath203 can be decreased but the oscillations in the relative wave functions demand an increasingly big number of points .
a typical grid for the case @xmath204 is g@xmath205 we have shown in table [ dephan_nnn ] the phase - shifts for different @xmath206 channels as a function of the center of mass energy .
for all of them , we have arbitrarily chosen , as in ref .
@xcite , the determination @xmath207 . for the s=0 case the comparison between grids t@xmath197 and
t@xmath199 has been made , showing a good stability despite the fact that grid t@xmath197 gives only a poor description of the asymptotic state . as it has been already emphasized in section 2 , the key point in our approach to the scattering problem
lies in the coherence between the asymptotic state and the numerical solution of the 4n problem , rather than in a very precise description of it .
the results corresponding to grid t@xmath199 are considered as converged .
a zero energy calculation directly provides us with the scattering length .
the results , given in table [ tab_sl ] , show a high stability with respect to grid variations ( t@xmath199 and t@xmath201 ) and our estimated accuracy is given in column 3 .
these values are in agreement with the existing published calculations @xcite . for @xmath208
they are close to those obtained in @xcite although with the original potential parameters .
the remaining low energy parameters have been extracted from the phase - shifts and are given in table [ tab_lep ] for the different @xmath206 channels .
they are defined in the effective range expansion : @xmath209\ ] ] where @xmath210 in the usual case but has to be modified in the presence of a near threshold singularity . according to @xcite ,
we have taken the form ( @xmath211 real ) : @xmath212 the validity of the expansion ( [ mera ] ) in the energy region below the first inelastic threshold is displayed in figure [ fig_kcotd ] .
the usual effective range approximation , i.e. @xmath210 and the @xmath213 contribution neglected in ( [ mera ] ) , works very well in the whole energy domain for all but the @xmath214 channel , in which the existence of a near threshold @xmath3he excited state requires the explicit inclusion of the pole contribution ( [ pole ] ) .
it is worth noticing that the contributions coming from the @xmath215 term are very small and have been included only for completeness . on the contrary
the pole contribution , existing only in the @xmath214 channel , is essential .
expansion ( [ mera ] ) provides a very accurate parametrization of the s - wave scattering amplitude in all the energy region below the break - up threshold .
we would like to emphasize here the coherence between the @xmath3he@xmath4 and the scattering results in the @xmath214 channel .
inserting the effective range expansion ( [ mera ] ) in the s - wave scattering amplitude @xmath216 produces an imaginary pole in the upper complex q - plane with value @xmath217 .
the corresponding energy @xmath218 mev is in close agreement with the direct calculation given in table [ tab_bs ] .
a step beyond the s - wave approximation has been taken by keeping the s - wave interaction alone but allowing non zero values in the angular momentum expansion of the fy amplitudes .
table [ ldifl ] shows the sensibility of the scattering lengths values when the maxima of @xmath219 and @xmath77 vary from 0 to 2 . in the particular interaction model we are considering
, the number of channels describing the 3n asymptotic state remains the same whereas the number of channels @xmath220 of the 4n problem considerably increases .
these results have been obtained with the grid t@xmath199 and @xmath185 completed with a suitable z - grid .
the values are well converged with @xmath221 except in the @xmath214 case , where the big value of the scattering length makes this state rather sensitive to small parameter variations .
of particular interest is to apply the preceding results to the description of n+t cross section . on the one hand
this is a pure @xmath208 channel , free from the difficulties related to the coulomb interactions . on the other hand accurate low energy scattering data exist @xcite , showing a structure at neutron laboratory energy @xmath222 mev which is supposed to be created by a series of p - waves resonances @xcite .
the calculations discussed in the preceding section have thus been completed up to the first 3-body break - up threshold by the inclusion of the first negative parity states @xmath223 corresponding to n+t relative p - waves .
the resulting total cross sections are plotted in figure [ nt1 ] .
the contributions from n+t relative s- and p - waves are distinguished .
we notice that the mt i - iii model conserves separately l and s and consequently the @xmath223 states coming from @xmath224 are degenerate .
the corresponding cross sections differ only by statistical factors .
the remaining @xmath225 state comes from an @xmath226 coupling . in view of these results ,
several remarks are in order : 1 .
the scattering lengths obtained in the s - wave approximation ( table [ tab_sl ] ) gives a slightly overestimated value for the zero energy cross section @xmath227 b. the experimental extrapolated zero energy cross section is @xmath228 b @xcite .
however the inclusion of higher partial waves in the fy expansion ( table [ ldifl ] ) significantly reduces the s - wave approximation result to @xmath229 b in closer agreement with experiment .
the comparison of the separated spin contribution is not possible since the values of the spin - dependent scattering lengths , summarized in table [ tab3 ] , are still controversial .
despite the simplicity of this model , the agreement with experimental data in the resonance region , is very good .
the n+t p - waves resonances are generated by an nn s - wave interaction alone .
an effective 1 + 3 p - wave potential is created due to the exchange mechanism between the four nucleon .
we remark however that a first attempt to describe this cross section with @xmath230 in the k components but keeping zero all the remaining angular momenta in ( [ kpw ] ) failed .
the incoming neutron seems thus to be more sensible to the virtual excitations of the triton than to the nn pair interactions themselves .
we have calculated the contribution coming from the n+t relative d - waves .
they are given by the positive parity states @xmath231 , @xmath232 .
the corresponding phase - shifts at 6 mev c.m .
kinetic energy are @xmath233 and @xmath234 , which contributes only a few mb to the total cross section .
the results displayed figure [ nt1 ] can so be considered as fully converged in the mt i - iii model .
the last point to be presented in this work concerns the reaction @xmath235 in the isospin approximation employed throughout this work , both n+@xmath1he and p+t thresholds are degenerate and correspond to the n+nnn of the preceding section .
the pauli principle imposes for the deuteron - deuteron channel @xmath236 .
if we assume the final state d+d to be in a relative s - wave one has @xmath237 and @xmath238 .
the @xmath239 state requires a relative angular momentum @xmath240 in the initial n+nnn channel and it is expected to give small contribution at very low d+d energy .
we will consider only the @xmath241 state .
the s - matrix is defined according to ( [ smatrix ] ) .
we present in table [ smat ] the s - matrix elements at several energies , chosen with respect to the nn+nn threshold .
n+nnn is referred to as channel 1 , nn+nn as channel 2 .
the symmetry and unitarity properties are there fulfilled at the level of @xmath151 , what corresponds to the accuracy of our results . with the conventions used above , the scattering cross sections in a given partial wave ( @xmath242 ) in presence of several channels @xmath243 are given by : @xmath244 where the notation @xmath245 holds for @xmath246 .
@xmath247 denotes the spin of the colliding cluster @xmath248 .
the corresponding values with @xmath249 n+nnn , nn+nn are given in figure [ siginel ] ( filled circles ) .
we remark no accident in the n+nnn cross section when the inelastic nn+nn threshold is open .
the nn+nn @xmath5 n+nnn cross section displays the usual @xmath250 law of the inelastic process and the crossed channel n+nnn @xmath5 nn+nn the expected @xmath251 law .
the elastic d+d cross section has been calculated neglecting coulomb interactions which will make this quantity divergent . however , due to the absence of nearthreshold singularity , one can expect small corrections to the low energy parameters obtained in this way , like in the n+n versus p+p case . of particular interest
is the extraction of the imaginary part of the strong d+d scattering length , a quantity which controls the fusion rate in the process d+d
@xmath5 n+@xmath1he .
this can be done either from the linear behavior of the d+d phase - shifts @xmath252 ( with @xmath253 ) , as displayed in fig .
[ ddd1 ] , or from the inelasticity in the non diagonal s - matrix element which behaves like @xmath254 .
both methods agree with high accuracy and lead to the following values : @xmath255 we remark a very small value of @xmath256 that should be only slightly modified once the coulomb interaction is switched on .
this small value is due to the small overlapping between the k and h configurations which respectively govern the n+nnn and nn+nn asymptotic states .
we have presented the first solution of faddeev - yakubovsky equations in configuration space for the scattering states in the 4n system .
they concern both the n+nnn elastic scattering and its coupling to the first inelastic nn+nn channel .
the n+nnn elastic phase - shifts have been calculated for the different spin and isospin channels .
the low energy parameters have been extracted and the validity of the effective range approximation in the energy region below the 3-body break - up threshold has been analyzed .
we have in particular found the importance of including the pole contribution in the @xmath214 case due to the vicinity of the first @xmath186 excitation .
we have also found that in the framework of our model this @xmath2 state is bound at 0.25 mev below the 3n threshold .
the study of the two - body correlation functions showed that the structure of this state is a 1 + 3 configuration rather than a breathing mode , as it is usually accepted .
the coherence between the binding energy of this state and the scattering results has been emphasized .
the n+t scattering cross section has been treated with a special interest and the first negative parity states have been included , to account for the structure experimentally observed .
the elastic cross section is well described by the simple mt i - iii interaction , especially in the resonance region .
the n+t p - waves resonance is thus reproduced by a nn pure s - wave interaction .
this shows the difficulty to understand this structure in terms of the nn interaction alone .
it is created by the direct and exchange mechanism between the incoming neutron and the target nucleons .
the s - matrix coupling the n+nnn and nn+nn channels has been obtained as well as the corresponding cross sections .
the analysis of these results allows the extraction of the nn+nn scattering length , whose imaginary part controls the fusion process
d+d @xmath5 n+@xmath1he .
its value turned out to be very small ( @xmath257 fm ) .
we are indebted to pr . c. gignoux for his continuous teaching and helpful discussions during the elaboration of this work .
the numerical calculations were initiated on the cray - t3d / t3e at the cgcv ( centre grenoblois de calcul vectoriel , cea ) , and completed on the cray - t3e at the idris ( institut du dveloppement et des ressources en informatique scientifique , cnrs ) .
we are grateful to the staff members of these two organizations for their kind hospitality and useful advice .
this appendix aims at the complete expression of the functions @xmath258 appearing in ( [ fse ] ) in terms of the corresponding quantum numbers @xmath37 and @xmath189 . by projection onto the appropriate k and h basis , the r.h.s . of the equations ( 11 ) gives raise to integral terms as shown in ( 13 ) , involving some very complicated functions @xmath259 , @xmath260 , @xmath261 , generated by permutations operators .
the permutation operators @xmath45 are completely defined by their action upon each ket of a complete basis .
this basis is chosen to be an ordered quadruplet @xmath262 , where the given @xmath263 value represents the state of the @xmath263 particle , including the space , spin and isospin degrees of freedom .
this basis is a generalization of the one given in ( [ baser ] ) .
it is assumed that @xmath45 corresponds to an exchange between the @xmath263 and @xmath264 set of quantum numbers , e.g. @xmath265 more complex permutations are then transparent , e.g. @xmath266 the way the permutation operators act upon the coupled bases ( k or h ) can be deduced from this basic feature . in the configuration space ,
e.g. , a single operator results in a generalized rotation of the jacobi coordinates . note that because of the symmetry of the bases , the contribution of the operators @xmath267 , @xmath268 and @xmath269 are identical with those of @xmath270 , @xmath271 and @xmath272 .
all the permutation operations can be seen as , at most , two successive rotations , each of them involving only two coordinates , e.g. first of all @xmath274 and @xmath81 , then @xmath191 and one of the preceding rotated coordinates .
so it is convenient to use the functions appearing in the 3-body problem @xcite , defined as follows .
suppose we are to calculate the projection of a given expression @xmath275_l f(x',y')$ ] , @xmath276 being an arbitrary function of @xmath277 et @xmath278 , onto a bipolar - harmonic basis @xmath279_l$ ] , where the following relation holds : @xmath280 we define @xmath273 functions such that : @xmath281^*_l \cdot [ y_{l_x'}(\hat{x } ' ) y_{l_y'}(\hat{y}')]_l \cdot f(x',y ' ) = \displaystyle \frac{1}{2 } \int_{-1}^1 du \cdot { \cal{h}}^l_{l_x l_y , l_x ' l_y'}(x , y , u ) \cdot f(x',y')\ ] ] where , in the integral term , @xmath277 et @xmath278 are obtained from ( [ matrotxy ] ) with the constraint @xmath282 .
these functions are given for example by : @xmath283 where @xmath284 et @xmath285 stands for the @xmath286-legendre polynomial . obviously , in the four - body case
, one has to deal with cumbersome recoupling , to isolate bipolar harmonics from the more complex coupling scheme ( [ devcoupling ] ) .
nevertheless all the work can be built upon this basic functions .
@xmath287 is non - zero when both @xmath37 and @xmath189 are of h type .
let us define then @xmath288 , @xmath289 , @xmath290 by @xmath291 @xmath287 is given by : @xmath292 where the isospin contribution @xmath293 is @xmath294 .
it is necessary to define in this case @xmath308 , @xmath309 , @xmath310 and an intermediate coordinate @xmath311 such that : @xmath312 the corresponding expression for @xmath313 is : @xmath314 with the constraints @xmath282 and @xmath315 , and where : @xmath316 we define again @xmath308 , @xmath309 , @xmath310 and an intermediate coordinate @xmath311 in the following way : @xmath317 the corresponding expression for @xmath313 is now : @xmath318 where @xmath319 and with the constraints @xmath282 and @xmath315 .
ccccccccccc + k & @xmath320 & @xmath321 & @xmath322 & @xmath323 & @xmath324 & @xmath219 & @xmath325 & @xmath326 & @xmath77 & @xmath327 + @xmath5 & 1&1/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + @xmath5 & 0&1/2&0&1&1&0&1/2&1/2@xmath328&0&1/2 + h & @xmath320 & @xmath329 & @xmath322 & @xmath323 & @xmath330 & @xmath219 & @xmath331 & @xmath332 & @xmath333 & @xmath77 + & 1&1&0&0&0@xmath328&0&0&0@xmath328&0&0 + @xmath334 & 0&0&0&1&1@xmath328&0&1&1@xmath328&0&0 + + k & @xmath320 & @xmath321 & @xmath322 & @xmath323 & @xmath324 & @xmath219 & @xmath325 & @xmath326 & @xmath77 & @xmath327 + @xmath5 & 1&1/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + @xmath5 & 0&1/2&0&1&1&0&1/2&1/2@xmath328&0&1/2 + & 0&1/2&0&1&1&0&1/2&3/2@xmath328&0&1/2 + h & @xmath320 & @xmath329 & @xmath322 & @xmath323 & @xmath330 & @xmath219 & @xmath331 & @xmath332 & @xmath333 & @xmath77 + & 0&0&0&1&1@xmath328&0&1&1@xmath328&1&0 + + k & @xmath320 & @xmath321 & @xmath322 & @xmath323 & @xmath324 & @xmath219 & @xmath325 & @xmath326 & @xmath77 & @xmath327 + @xmath5 & 1&1/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + @xmath5 & 0&1/2&0&1&1&0&1/2&1/2@xmath328&0&1/2 + & 1&3/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + h & @xmath320 & @xmath329 & @xmath322 & @xmath323 & @xmath330 & @xmath219 & @xmath331 & @xmath332 & @xmath333 & @xmath77 + & 1&1&0&0&0@xmath328&0&0&0@xmath328&0&0 + + k & @xmath320 & @xmath321 & @xmath322 & @xmath323 & @xmath324 & @xmath219 & @xmath325 & @xmath326 & @xmath77 & @xmath327 + @xmath5 & 1&1/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + @xmath5 & 0&1/2&0&1&1&0&1/2&1/2@xmath328&0&1/2 + & 1&3/2&0&0&0&0&1/2&1/2@xmath328&0&1/2 + & 0&1/2&0&1&1&0&1/2&3/2@xmath328&0&1/2 + h & @xmath320 & @xmath329 & @xmath322 & @xmath323 & @xmath330 & @xmath219 & @xmath331 & @xmath332 & @xmath333 & @xmath77 + & 1&0&0&0&0@xmath328&0&1&1@xmath328&1&0 + & 0&1&0&1&1@xmath328&0&0&0@xmath328&1&0 + ccdddcdddcddd + x & & 20 & 1.30 & 10.0 & & - & - & - & & - & - & - + y & & 15 & 1.25 & 12.0 & & - & - & - & & - & - & - + z & & 15 & 1.25 & 15.0 & & - & - & - & & - & - & - + + x & & 15 & 1.30 & 10.0 & & - & - & - & & - & - & - + y & & 10 & 1.25 & 12.0 & & - & - & - & & - & - & - + z & & 10 & 1.25 & 15.0 & & - & - & - & & - & - & - + + x & & 08 & 1.30 & 08.0 & & 05 & 1.10 & 20.0 & & - & - & - + y & & 07 & 1.30 & 10.0 & & 05 & 1.10 & 30.0 & & - & - & - + z & & 07 & 1.20 & 10.0 & & 13 & 1.10 & 80.0 & & 10 & 1.00 & 150.0 + clllddd @xmath335 & @xmath322 & @xmath219 & @xmath77 & @xmath336 & @xmath337 & @xmath338 + 2 + 2 & 0 & 0 & 0 & 30.302 & 8.769 & 0.257 + 8 + 2 & 0 & 0,1 & 0,1 & 30.319 & 8.763 & 0.250 + 16 + 3 & 0 & 0,1,2 & 0,1,2 & 30.324 & 8.770 & 0.257 + ccdddcddd + x & & 08 & 1.30 & 14.0 & & 01 & - & 18.0 + y & & 07 & 1.20 & 19.0 & & 02 & 1.00 & 29.0 + + x & & 12 & 1.30 & 14.0 & & 01 & - & 18.0 + y & & 10 & 1.20 & 19.0 & & 02 & 1.00 & 29.0 + + x & & 18 & 1.20 & 14.0 & & 02 & 1.00 & 18.0 + y & & 15 & 1.10 & 19.0 & & 04 & 1.00 & 29.0 + ddd @xmath339 ( mev ) & + & t@xmath197 & t@xmath199 + 0.05 & 169.97&169.95 + 0.1 & 165.85&- + 0.5 & 148.99&- + 1.0 & 137.13&137.07 + 2.0 & 121.79&- + 3.0 & 111.27&- + 4.0 & 102.93&102.82 + 5.0 & 96.18 & 96.13 + 6.0 & 90.18 & 90.24 + & t@xmath197 & t@xmath199 + 0.01&165.17&164.05 + 0.02&159.17&157.66 + 0.05&147.73&- + 0.07&142.29&- + 0.1 & 135.72&132.98 + 0.2 & 120.52&- + 0.3 & 109.52 & 106.95 + 0.5 & 95.81 & - + 0.7 & 85.57 & - + 1.0 & 74.08 & 71.05 + 2.0 & 49.92 & 47.16 + 3.0 & 34.71 & 32.07 + 4.0 & 23.60 & 21.04 + & & t@xmath199 + 0.03&&172.97 + 0.06&&170.07 + 0.12&&166.0 + 0.3 & & 158.02 + 0.9 & & 142.84 + 1.8 & & 129.11 + 2.7 & & 119.43 + 3.6 & & 111.84 + 5.4 & & 100.10 + 6.3 & & 94.84 + & & t@xmath199 + 0.01&&176.46 + 0.02&&174.99 + 0.05&&172.09 + 0.1 & & 168.83 + 0.5 & & 155.23 + 1.0 & & 145.35 + 2.0 & & 132.06 + 3.0 & & 122.53 + 4.0 & & 114.96 + 5.0 & & 108.68 + 6.0 & & 103.24 + ccdddcd + s & t & triton 8&triton 10 & + 0 & 1 & 4.13 & 4.13 & 4.13 & @xmath340 & 0.01 + 1 & 1 & 3.73 & 3.73 & 3.73 & @xmath340 & 0.01 + 0 & 0 & 14.78 & 14.76 & 14.76 & @xmath340 & 0.02 + 1 & 0 & 3.25 & 3.25 & 3.25 & @xmath340 & 0.01 + ccddddd s&t & @xmath84 ( fm ) & @xmath341 ( fm ) & @xmath215 ( fm@xmath1 ) & @xmath211 ( @xmath343 & @xmath84 @xcite + 0&1 & 14.75 & 6.75 & 0.462 & - & + 1&1 & 3.25 & 1.82 & 0.231 & - & + 0&0 & 4.13 & 2.01 & 0.308 & 0.505 & 4.0 + 1&0 & 3.73 & 1.87 & @xmath344 0 & - & 3.6 + ddcccc & + n+nnn&nn+nn&@xmath345&@xmath346&@xmath347&@xmath348 + 4.1253&0.05 & 0.772 + 0.634i & -0.00473 + 0.0472i & -0.00471 + 0.0470i & 0.882 - 0.468i + 4.1753&0.10 & 0.782 + 0.621i & 0.000488 + 0.0564i & 0.000486 + 0.0563i & 0.771 - 0.634i + 4.2753&0.20 & 0.802 + 0.594i & 0.0110 + 0.0661i & 0.0110 + 0.0661i & 0.566 - 0.821i + 4.3253&0.25 & 0.811 + 0.581i & 0.0160 + 0.0690i & 0.0161 + 0.0691i & 0.472 - 0.879i + 4.3753&0.30 & 0.820 + 0.567i & 0.0209 + 0.0711i & 0.0210 + 0.0713i & 0.382 - 0.921i + 4.5753&0.50 & 0.854 + 0.514i & 0.0391 + 0.0748i & 0.0393 + 0.0752i & 0.0659 - 0.994i + 5.0753&1.00 & 0.919 + 0.381i & 0.0759 + 0.0679i & 0.0758 + 0.0678i & -0.480 - 0.871i + 5.5753&1.50 & 0.961 + 0.253i & 0.103 + 0.0508i & 0.102 + 0.0506i & -0.784 - 0.610i + 6.0753&2.00 & 0.983 + 0.133i & 0.120 + 0.0292i & 0.123 + 0.0301i & -0.934 - 0.335i + dcddcdc & + 3.91 & @xmath340&0.12 & 3.6 & @xmath340&0.1 & @xcite + 4.98 & @xmath340&0.29 & 3.13 & @xmath340&0.11 & @xcite + 2.10 & @xmath340&0.31 & 4.05 & @xmath340&0.09 & @xcite + 4.453&@xmath340&0.10 & 3.325&@xmath340&0.016 & @xcite + \(a ) elastic scattering at zero - energy : @xmath349 asymptotically linear towards @xmath81 + + ( b ) elastic scattering with positive kinetic energy : @xmath349 asymptotically oscillating towards @xmath81 + | the faddeev - yakubowsky equations in configuration space have been solved for the four nucleon system .
the results with an s - wave interaction model in the isospin approximation are presented .
they concern the bound and scattering states below the first three - body threshold .
the elastic phase - shifts for the n+nnn reaction in different ( @xmath0 ) channels are given and the corresponding low energy expansions are discussed .
particular attention is payed to the n+t elastic cross section .
its resonant structure is well described in terms of a simple nn interaction .
first results concerning the s - matrix for the coupled n+nnn - nn+nn channels and the strong deuteron - deuteron scattering length are obtained . |
soon after the discovery of asymptotic freedom , it has become common knowledge that quantum chromodynamics ( qcd ) under extreme conditions , such as high temperature and/or density , is a theory of weakly coupled quarks and gluons @xcite .
this phase of qcd , called the quark - gluon plasma ( qgp ) , has attracted a considerable attention over the past few decades . theoretically , it is imperative to establish the thermodynamic properties of qcd at finite temperature and density in order to confront the experimental data of the heavy - ion collision .
the first principle lattice qcd calculations have revealed a great deal of information on the finite temperature phase transition(s ) between the hadronic phase and the qgp .
an important task would be to correctly interpret in concrete physical terms what is going on during the rapid transition . in qcd , at finite temperature
, it is well known that there are two different phase transitions in different parameter regions ; the chiral phase transition in theories with massless quarks and the deconfinement phase transition in theories with infinitely heavy quarks . since the parameters of the real world are rather close to the chiral limit , chiral symmetry is often emphasized in constructing phenomenological models and describing the phase transition .
one could argue that the deconfinement phase transition would have no definite meaning in this parameter region and could be forgotten altogether . yet
the word ` deconfinement ' is quite appealing .
the qgp is often referred to as a deconfined phase .
it is indeed tempting to interpret the rapid increase of the energy density observed in the lattice simulations as the liberation of the qcd degrees of freedom , or ` deconfinement ' .
perhaps the reason of adherence to deconfinement or the deconfinement phase transition is not merely an intuitive one .
since the early days of finite temperature qcd lattice simulations , it has been empirically known that the chiral phase transition and the deconfinement phase transition are tightly correlated although there seems to be no logical relation between them _ a priori _ @xcite . from these results
one expects that some remnants of the deconfinement phase transition might be taken over by the chiral phase transition . in this paper
, we argue that the above expectation is correct . the chiral phase transition and
the deconfinement phase transitions are just different limits of a single crossover phase transition .
the key observations are the glueball - sigma mixing and the level repulsion which ` lock ' the two phase transitions @xcite .
this helps dissolve the ambiguous issue of ` deconfinement ' and enables us to have a clearer picture of the nature of the finite temperature qcd phase transition .
the paper is organized as follows . in section
ii , we briefly review some basic facts about the chiral and deconfinement phase transitions and pose a ` puzzle ' long known in finite temperature lattice qcd simulations .
section iii contains our main results .
we present a possible scenario to explain the puzzle and argue that the phenomenon is a global feature of the qcd phase diagram in the ( @xmath0 ) plane ( @xmath1 is the temperature and @xmath2 is the current quark mass ) . in section iv , we speculate about qcd - like theories such as qcd with adjoint quarks and in the large @xmath3 limit .
section v is devoted to our conclusions .
the deconfinement phase transition is well defined only in theories with infinitely heavy quarks . above the critical temperature ,
the polyakov loop @xcite , @xmath4 where @xmath5 is the path ordering operator and @xmath6 are the su(@xmath7 generators in the fundamental representation , develops a nonzero expectation value which breaks center symmetry . in an su(3 ) pure gauge theory the transition is known to be first order @xcite .
when heavy quarks are added to the theory , the first order phase transition weakens .
it eventually ends in a second order phase transition point where @xmath8 becomes continuous .
the universality class of this phase transition is that of the three dimensional ising model . the critical quark mass at is estimated to be @xmath9 gev in the two - flavor lattice simulations @xcite .
in contrast , chiral symmetry is a true symmetry only in theories with massless quarks .
the order parameter of the chiral phase transition is the quark condensate @xmath10 . for three flavors ,
the phase transition is first order @xcite . at finite but small quark mass ( which we take to be equal for all flavors ) , the first order transition persists and eventually ends in a second order phase transition point .
the jump in the chiral condensate disappears and the scalar meson field @xmath11 ( with certain flavor content ) becomes gapless at this point @xcite .
the location of and depends on the number of flavors and their mass ratios .
[ for example , lies on the @xmath12 axis for two flavors . ]
the argument in the next section does not depend on the details of the flavor sector , however . from these basic facts alone
, obviously there is no logical relation between the chiral and deconfinement phase transitions .
various possibilities have been argued in the literature @xcite .
a common argument is that , since confined vacuum necessarily breaks chiral symmetry @xcite , which suggests that the energy scale of the chiral symmetry breaking is higher than that of confinement , chiral symmetry restoration occurs at a somewhat higher temperature than deconfinement .
[ see , however , @xcite .
] this is in accord with our physical intuition ; it is easy to imagine massive constituent quarks but it is hard to imagine massless protons . if there were no lattice simulations at finite temperature , theorists would have drawn a phase diagram like fig . [ 2 ] , incorporating what is known from the universality argument .
naive phase diagram in the ( @xmath1 , @xmath2 ) plane .
the solid ( dotted ) lines denote the first order ( crossover ) phase transition . ]
however , the realistic phase diagram obtained from the lattice simulations , fig .
[ fig ] , is qualitatively different from our naive expectation @xcite .
first , the critical temperature of the deconfinement phase transition in the pure gauge theory @xmath13 mev is _ invariably higher _ than that of the chiral phase transition in the chiral limit @xmath14 mev .
[ one may think that it is meaningless to compare these two numbers because they are defined in completely different theories .
however , this is a suggestive result and can be naturally explained within our scenario . ]
the more striking fact is that the pseudocritical temperatures determined from the peak position of the susceptibility of each order parameter ( dotted curves in fig .
[ fig ] ) appear to coincide for _ all _ values of the quark mass , which means that two crossover curves are smoothly connected at an intermediate point .
it is as if the deconfinement phase transition in the pure gluonic theory is ` guided ' towards the chiral phase transition traveling over an infinite range of the quark mass .
this is a surprising result .
how can the deconfinement phase transition defined in a theory _ without _ quarks ` know ' about the chiral phase transition ?
realistic phase diagram observed in lattice simulations . ]
the interplay between the two phase transitions is an interesting problem .
first gocksch and ogilvie have observed that what is responsible for the breaking of center symmetry is ( the inverse of ) the _ constituent _ quark mass rather than the current quark mass @xcite .
this suggests that the polyakov loop and the sigma meson are closely coupled and the critical fluctuation of one field induces a singularity in the other field .
more recent works @xcite further confirmed this interplay .
however , as pointed out in @xcite and emphasized in @xcite , this argument alone is insufficient to explain the observed data . in fact , a stronger condition is needed to reproduce the phase diagram fig .
[ fig ] . in this section
, we will make a serious attempt to understand the physics behind fig .
we shall divide our argument into three steps .
although there is little doubt that the polyakov loop is the correct order parameter of the deconfinement phase transition in pure gauge theories , a perennial question about the physical meaning of the polyakov loop remains @xcite . in a word ,
the polyakov loop is a fictitious excitation in euclidean space .
it is true that , even in the presence of dynamical quarks , the behavior of the polyakov loop indicates a clear signal of the phase change .
however , it seems that the above conceptual barrier has long been an obstacle to our satisfactory understanding of the phase transition in the presence of dynamical quarks .
our first step is to propose an alternative characterization of the deconfinement phase transition .
specifically , we wish to show that the _ glueball _ can equivalently describe the critical phenomena at . for this purpose , first we work in the pure su(2 ) gauge theory . in this case
the deconfinement phase transition is second order @xcite whose universality class is that of the three dimensional ising model .
the string tension ( ` mass ' of the polyakov loop ) @xmath15 goes to zero smoothly as the critical temperature is approached @xmath16 where @xmath17 is the reduced temperature and the universal exponent is @xmath18 @xcite .
the infinite correlation of the polyakov loop entails a massless mode in physical excitations .
this can be seen by considering the specific heat @xmath19 .
near the critical point , the singular part of the specific heat may be represented by @xmath20 where @xmath21 and @xmath22 is the spatial dimension . of the ginzburg - landau model . ]
the subscript c denotes the connected part .
is real for su(2 ) and @xmath23 stays nonzero at any temperature . ] in the theory of critical phenomena , the original dimension of an operator has no meaning .
the polyakov loop is dimensionless , but it has nothing to do with the actual _ scaling dimension _ which enters in the calculation of correlation functions .
the scaling dimension of the order parameter squared is @xmath24 @xcite .
thus the explicit structure of the divergence is @xmath25 where we have assumed that the connected part of the correlator in ( [ c ] ) exponentially decays as @xmath26 at large distances . matching the powers of @xmath27 and using josephson s law @xmath28 ( hyperscaling ) , we obtain , @xmath29 hence the operator @xmath30 excites , among other states , at least one state with zero screening mass at the phase transition . by using an su(2 )
identity @xmath31 where @xmath32 is the polyakov loop in the adjoint representation ( i.e. , the generators @xmath33 in eq .
( [ 11 ] ) are replaced by those in the adjoint representation ) , or just by noticing that @xmath30 is invariant under the center transformation , @xmath34 , we see that the massless state is the @xmath35 glueball . should be called the ` gluelump ' instead of the glueball @xcite .
however , we expect that the difference ( if any ) dissolves at the deconfinement phase transition point . on the lattice
, the specific heat is measured from the integrated correlator of plaquettes which excite glueballs .
since the two expressions of the specific heat should give the correct singularity , the gluelump and the glueball must be the same excitation near the phase transition . ] eq . ( [ ttt ] ) is a trivial consequence of the scaling property at general second order phase transition points ; the only relevant scale is the correlation length @xmath36 .
nevertheless , in the present case it is a nontrivial relation ensuring the vanishing of the mass gap in physical excitations as opposed to the polyakov loop which has no particle interpretation in minkowski space - time .
note that although the polyakov loop and the glueball become massless with the same critical exponent @xmath37 , their susceptibilities diverge with different critical exponents ; @xmath38 for the glueball and @xmath39 for the polyakov loop ( the order parameter ) .
[ compare with the su(3 ) case below . ] a significant decrease of the @xmath35 glueball screening mass near the su(2 ) deconfinement transition has been observed on the lattice @xcite . + one might suspect that the vanishing of the glueball mass would contradict with the presence of the trace anomaly . in the simple phenomenological potential of glueballs at zero temperature @xcite , the glueballs acquire their mass entirely from the trace anomaly , or the gluon condensate .
@xmath40 since we do expect that the trace anomaly persists at any temperature , this appears to contradict with our assertion . however , the foundation of this objection is obscure .
first , the exact lattice sum rule for the glueball mass @xcite shows that only @xmath41 of the glueball mass comes from the expectation value of the trace anomaly in the glueball state measured _ relative _ to the vacuum .
although it might be useful for certain purposes , eq .
( [ glu ] ) can not be taken seriously in our context . secondly , as observed in @xcite , the symmetry organizing the glueball spectrum changes near the critical temperature of both the su(2 ) and su(3 ) pure gauge theories .
at low temperatures , the zero temperature symmetry is dynamically realized in the glueball spectrum @xcite . near the phase transition ,
the zero temperature symmetry is superseded by the finite temperature one where the inclusion of the timelike links in interpolating operators makes a drastic change in the screening masses .
the glueballs which become massless at the phase transition are the ` electric ' glueballs ( those containing @xmath42 , or time - like links ) , while the ` magnetic ' glueballs ( composed of @xmath43 , or space - like links ) remain massive .
this crucial difference has been overlooked in most phenomenological approaches .
[ see , however , @xcite .
] finally , but most cogently , eq .
( [ ttt ] ) is true for any second order phase transition .
if all the glueballs were massive at the phase transition , the specific heat would be finite .
, the correct singularity in the specific heat has not been confirmed on the lattice so far .
see , however , @xcite . ]
this is a catastrophe because it means that the apparent singularity in the polyakov loop has nothing to do with thermodynamics and there is no ( second order ) phase transition .
+ let us go back to the end - point of su(3 ) qcd with dynamical quarks .
since center symmetry is explicitly broken by dynamical quarks , the polyakov loop acquires a nonzero expectation value @xmath44 at .
thus the phase transition at can be characterized by the infinite correlation length ( or the vanishing screening mass ) of the shifted polyakov loop @xmath45 using @xmath46 as an order parameter , one can construct a ginzburg - landau theory near and discuss the singular behavior of various susceptibilities .
however , due to our identification @xmath47 the @xmath48 glueball field @xmath49 contains a linear term in @xmath46 .
hence the screening mass of the polyakov loop which vanishes at is equal to the screening mass of the glueball , and we may as well construct a ginzburg - landau theory of @xmath50 around . note that this is a direct consequence of the explicit breaking of center symmetry .
unlike the su(2 ) case , the glueball qualifies as an order parameter field ; its susceptibility diverges with the exponent @xmath51 .
the two descriptions of the critical phenomena at , with the polyakov loop or with the glueball , are equivalent . in the presence of dynamical quarks ,
the singular behavior of the polyakov loop represents the character change of the glueball .
since the latter admits a direct physical interpretation , we will exclusively focus on the glueball in the following .
the next step is a rather trivial , but important observation .
_ in the presence of dynamical quarks , the glueballs must mix with the @xmath52 states_. in fact , the correct massless mode associated with the second order phase transition at is a linear combination of the pure glueball field @xmath50 and the scalar meson field @xmath53 with large glueball content , @xmath54 the other linear combination with large sigma meson content , @xmath55 is massive , hence decouples at . the mixing eq .
( [ g ] ) is physically obvious , and it is again a consequence of the explicit breaking of center symmetry . in the language of critical phenomena , the ginzburg - landau potantial has the following expansion , @xmath56g'+o(g'^2)+o(g'^3)+o(g'^4 ) , \end{aligned}\ ] ] where ( @xmath57 , @xmath58 ) is the location of , and @xmath59 and @xmath60 are smooth functions of @xmath1 and @xmath2 .
@xmath61 is an interpolating field of the glueball whose expectation value at is subtracted . in powers of @xmath46 .
the coefficient of the linear term is unchanged up to a multiplicative constant . ]
the equation of the tangential line at is given by @xmath62 which are in general not parallel ( @xmath63 ) to the mass axis in the absence of symmetry .
the linear dependence on @xmath2 in the coefficient of @xmath46 makes the chiral susceptibility @xmath64 divergent with the same exponent @xmath39 as the polyakov loop susceptibility , the glueball field susceptibility and the specific heat .
is very large , the mixing is weak and it would be difficult to observe the singularity of @xmath65 in numerical simulations @xcite . ] in the same way , near the glueball field mixes with the sigma meson field which is critical . the simultaneous enhancement of the chiral susceptibility and the polyakov loop susceptibility along _ each _ crossover phase transition line is a natural consequence of the mixing @xcite .
the coincidence of the peak positions is trivial .
it is not a puzzle .
it is just an evidence of the glueball - sigma mixing .
the real puzzle of the lattice data is that there is only _
one _ such coincidence in the @xmath66 plane .
namely , the mixing eq .
( [ g ] ) alone does not exclude the phase diagram fig . [ 2 ] in which the coincidence occurs _ twice _
, first at the deconfinement phase transition and then at the chiral phase transition .
so what is going on in fig .
[ fig ] ?
the last step of our argument is a possible bridge between the two phase transitions .
we have defined the @xmath67 field eq .
( [ g ] ) as the correct massless field at the end - point of the deconfinement phase transition .
on the other hand , since the massive @xmath68 field has a large meson component there , naively we expect that it becomes critical ( massless ) at the end - point of the chiral phase transition .
this is what has been implicitly assumed to occur . however , this innocent - looking assumption is the principal cause of our confusion .
the two phase transitions are caused by _
different _ fields .
why do the two crossover lines meet ?
the crucial difference between fig .
[ 2 ] and fig .
[ fig ] is that in fig .
[ fig ] , _ both phase transitions are caused by the @xmath67 field_. since the critical field at should be predominantly the sigma meson , this is possible only if the mixing angle changes from @xmath69 to @xmath70 in an intermediate region of the quark mass .
such a continuous variation of the mixing angle is typical of a _ level repulsion _ in quantum mechanics .
thus we have reached a novel scenario of the finite temperature qcd phase transition : _ due to a level repulsion between the @xmath67 and the @xmath68 fields , the @xmath67 field continues to be the lightest field for all values of the quark mass .
simultaneous divergences and peaks in various susceptibilities are simply caused by the dropping of the @xmath67 field screening mass .
_ this gives a natural explanation to the otherwise puzzling lattice result , fig .
note that the somewhat counterintuitive ordering 280 mev @xmath71 170 mev mentioned in section ii is also a natural consequence of our scenario .
generally , a crossover line extends to the direction opposite to the symmetry broken phase .
if the phase transition is second order in the absense of external fields , even the equation of the crossover line is determined by the universality class up to a multiplicative constant .
the point is that center symmetry is broken in the _ high _ temperature phase .
this is a quite unusual phenomenon but is true for su(@xmath3 ) pure gauge theories .
thus , if the two phase crossover lines are smoothly connected by our scenario , the critical temperature of the deconfinement phase transition in the pure gauge theory must be larger than that of the chiral phase transition in the chiral limit .
+ we now argue that our level repulsion scenario is not an _ ad hoc _ idea just to explain the observed data , but rather a global feature of the qcd phase diagram in the ( @xmath1 , @xmath2 ) plane .
consider _ zero _ temperature qcd .
for small values of the quark mass , the sigma meson is a broad resonance whose mass is about 600 mev . on the other hand ,
the mass of the lowest glueball state ( which mixes with nearby quarkonium states ) is about 1.6 gev .
if the quark mass is sent to infinity , the sigma meson mass goes to infinity , while the glueball mass stays more or less the same .
naively , this implies that there is a level crossing between the two states at an intermediate value , roughly half the glueball mass @xmath72 mev , of the quark mass .
however , in a quantum mechanical system , quite generally levels with the same quantum numbers can not cross unless certain matrix elements accidentally vanish . therefore , what actually takes place is a level repulsion .
under this assumption at @xmath73 ( which seems to be reasonable to us ) , the level repulsion would continue to take place in the ( @xmath1 , @xmath2 ) plane up to the crossover phase transition and our scenario would be realized .
field were the order parameter at , there must have been a level crossing between the @xmath67 and @xmath68 fields . generally , in a two - parameter space ( @xmath1 , @xmath2 ) , a level crossing occurs along a line simply because of a geometrical reason @xcite .
therefore , this ( conventional ) scenario would lead to a level crossing at zero temperature , which is again unlikely . ]
thus the glueball - sigma mixing and the level repulsion provide not only a unified description of the chiral and deconfinement phase transitions , but also a consistent characterization of the entire qcd phase diagram in the ( @xmath1 , @xmath2 ) plane .
[ phase ] is a schematic plot of the lowest scalar screening mass below the crossover phase transition in our scenario .
the two phase transitions are indistinguishable . roughly speaking , they are the same phase transition . at
about @xmath74 800 mev , there is a smooth change of the degrees of freedom , i.e. , mesons at smaller @xmath2 to glueballs at larger @xmath2 , in the low energy effective theory of qcd at _ any _ temperature at least up to the crossover phase transition . in section iv , we will see another possible example of the level repulsion .
+ the lowest screening mass @xmath75 [ gev ] in the scalar channel below the crossover phase transition . at small @xmath2 ,
the scalar channel is dominated by the sigma meson @xmath53 . at large @xmath2
, it is dominated by the glueball @xmath50 . ]
basically our argument does not change at nonzero baryon / isospin chemical potential . the interplay between the sigma and the glueball ( or the polyakov loop ) continues to exist @xcite .
the only new feature is that as the qcd critical point is approached , the baryon number density begins to mix with the scalar mode and the baryon number susceptibility diverges with the same critical exponent as the chiral susceptibility @xcite .
lattice simulations @xcite as well as model calculations @xcite have shown that , in qcd with adjoint quarks , the chiral phase transition occurs at a distinctly higher temperature than the deconfinement phase transition . here
we analyze this problem from our point of view . for simplicity , we work in @xmath76 .
since adjoint quarks do not break center symmetry , the second order deconfinement phase transition persists for all values of the quark mass .
the effect of the adjoint quarks is simply to modify the coefficients of the ginzburg - landau potential @xmath77 . near a critical point ( @xmath78 , @xmath79 ) on the critical line
, @xmath77 can be expanded as @xmath80l^2+cl^4,\label{omega}\end{aligned}\ ] ] where @xmath59 , @xmath60 and @xmath81 are smooth functions of @xmath1 and @xmath2 . from ( [ omega ] ) we see that both the chiral susceptibility and the specific heat diverge with the critical exponent @xmath82 .
this means that although the glueball and the sigma meson are massless on the critical line , they are distinct from the polyakov loop whose susceptibility diverges with the exponent @xmath51 .
the mixing between the polyakov loop and the sigma meson is prohibited by center symmetry . on the other hand , at the chiral phase transition the chiral susceptibility diverges with the exponent @xmath51 . may be different from that at the deconfinement phase transition depending on the universality class . ] clearly , the two phase transitions can not be simultaneous . interestingly , even in this case there is a level repulsion ( or a level crossing ) between the two phase transitions .
since the critical line is almost parallel to the quark mass axis ( @xmath83 ) for small @xmath2 , the amplitude of the singular part of the chiral susceptibility is small .
this means that the massless mode responsible for the @xmath82-divergence is predominantly the glueball . on the other hand ,
the massless mode at the chiral phase transition is predominantly the sigma meson .
naturally , we expect that there is a level repulsion between the sigma meson and the glueball at an intermediate temperature
. it is intriguing to consider the phase diagram for different values of @xmath3 than three and in the limit of infinite @xmath3 @xcite .
lattice simulations have shown that , as @xmath3 increases , the first order deconfinement phase transition strengthens @xcite .
this means that a smaller value of @xmath2 is needed to wash out the first order phase transition , i.e. , the point moves downward .
since there does not seem to be anything special about three colors in our scenario , we expect that the point the point at some value of @xmath85 and the high and low temperature phases are separated by a single first order phase transition line for @xmath86
. however , the limit @xmath87 might be a different story .
the critical temperature of the deconfinement phase transition is expected to be of the order of @xmath88 , while that of the chiral phase transition is controversial . if we regard @xmath89 as the chiral symmetry breaking scale ( @xmath90 is the pion decay constant ) , it is of the order of @xmath91 .
our scenario appears to be incompatible with this scenario . on the other hand ,
there are arguments that the chiral phase transition temperature can not grow with @xmath3 ; it is of the order of the @xmath92 meson mass @xmath93 .
it is not our intention to go into the details of these arguments which have been considered by many people previously .
here we just point out that at least our scenario breaks down in the large-@xmath3 limit .
this is because , in the usual power counting rule of large-@xmath3 qcd , the glueball - meson mixing strength vanishes like @xmath94 .
in the absence of the mixing , the level repulsion is impossible and the ` unlocking ' of the two phase transitions could occur . from our viewpoint
, @xmath76 is special .
the deconfinement phase transition in the pure gauge theory is second order .
the ginzburg - landau theory of @xmath50 does not make sense at . nevertheless , the mixing between the polyakov loop and the glueball is possible when the dynamical quarks are added .
another notable feature is that the phase transition on the ` quark side ' is driven by the @xmath95 diquark ( ` baryon ' ) at finite temperature and density .
it is thus interesting to pursue a similar link ( glueball - sigma - diquark mixing ) in the phase diagrams of two - color qcd @xcite .
based on the existing lattice data , we have proposed a novel scenario of the finite temperature qcd phase transition . the chiral phase transition at zero quark mass and the deconfinement phase transition at infinite quark mass are continuously connected by the glueball - sigma mixing and the level repulsion .
this gives a precise meaning to the common but somewhat ambiguous statement : in qcd at high temperature , quarks are deconfined and chiral symmetry is restored . " we have also argued that our scenario can be naturally embedded in the entire phase diagram in the ( @xmath1 , @xmath2 ) plane ( fig . [ phase ] ) . at
any given value of the quark mass , the minimum of the @xmath67 field screening mass marks the well - defined onset of the qgp phase .
+ although we believe that our scenario is the correct interpretation of the finite temperature lattice simulations , direct evidences are still lacking .
the most convincing evidence would be obtained by the lattice qcd measurements of the glueball , the polyakov loop and the sigma meson screening masses along the crossover transition line . because of the mixing eq .
( [ g ] ) , they are all equal along the line . in principle , the mixing between the sigma and the glueball could occur in the entire phase diagram ( fig .
[ fig ] ) .
however , we expect that in most regions the mixing is too small to be observed .
for example , well above the chiral phase transition temperature , for small @xmath2 , the sigma meson and the pions fall in a multiplet due to the restored chiral symmetry .
the glueball mass is indeed different from the sigma mass in this region @xcite .
also at low temperatures , for small @xmath2 , the glueball will primarily mix with nearby scalar quarkonium states . as argued in section iii , the strong mixing
is expected to occur at rather large values of @xmath2 , say , @xmath96 mev .
detailed analyses of the lowest and the second lowest screening masses in this region would confirm or exclude our scenario .
for a deeper understanding of this phenomenon , it would be desirable to construct concrete models which embody the mixing mechanism . in models where the polyakov loop and the sigma are coupled in the effective potential ,
the pseudocritical temperatures of the two phase transitions are more or less close to each other @xcite .
however , it has been so far not successful to reproduce the complete ` locking ' of the two phase transitions .
the obtained phase diagram is qualitatively similar to fig .
[ 2 ] , which suggests that although these models capture the gross features of the phase transition , an essential coupling which reproduces the lattice data and demonstrates our scenario is still missing .
this appears to be a future theoretical challenge .
we thank s. r. beane , m. creutz , s. ejiri , l. mclerran , p. petreczky , r. d. pisarski , d. t. son , k. splittorff and m. a. stephanov for discussions and comments . y. h.
is supported by riken bnl research center and u.s .
department of energy grant # de - ac02 - 98ch10886 .
k. f. is supported by the japan society for the promotion of science for young scientists and supported in part by funds provided by the u.s .
department of energy under cooperative research agreement # df - fc02 - 94er40818 . | we argue that the chiral phase transition and the deconfinement phase transition in finite temperature qcd are just different limits of a single crossover phase transition driven by the lightest scalar field which is a mixture of the glueball and the sigma meson .
the key mechanism , the level repulsion between the sigma and the glueball , gives not only a unified description of the finite temperature phase transitions , but also a consistent characterization of the entire qcd phase diagram in the temperature - quark mass plane . |
pace missions can be impaired by intentional or unintentional jamming .
such a threat is particularly dangerous for telecommands ( tc ) , since the success of a mission may be compromised because of the denial of signal reception by the satellite .
it is well known that to counter the jamming threat , error correcting codes can be used jointly with direct sequence spread spectrum .
this topic has been investigated in previous literature , but rarely taking into account the peculiarities of the tc space link . as a consequence , in current standards or recommendations on tc space links , only weak countermeasures
are included , that do not appear adequate to face the increasing skill of malicious attacks .
whilst for the spreading technique a relevant advance is brought by the introduction of long cryptographic pseudo - noise sequences @xcite , the discussion is quite open as regards possible usage of new coding techniques . as a matter of fact
, the only error correcting code currently included in the standards and recommendations for tc applications @xcite , @xcite is the bch code with dimension @xmath0 and length @xmath1 exploiting hard - decision decoding .
the performance of this code in the presence of jamming is generally not good and it is possible to verify that significant losses appear even when an interleaver ( not currently included in the standard ) is employed .
actually , the performance of the bch@xmath2 code is unsatisfactory even when considered on the additive white gaussian noise ( awgn ) channel . for this reason a number of new proposals have been formulated with emphasis on binary and non - binary low - density parity - check ( ldpc ) codes @xcite , @xcite , @xcite . since good correction capability and short frames are needed , the most recent proposals consider codes with rate @xmath3 and @xmath4 , @xmath5 or @xmath6 .
using non - binary ldpc codes allows to improve on their binary counterparts @xcite : as an example , a non - binary ldpc@xmath7 code at a codeword error rate ( cer ) of about @xmath8 gains roughly @xmath9 db over the binary ldpc@xmath7 code .
the performance of these codes over jamming channels has not yet been investigated .
in this paper we consider different types of jamming signals , namely : pulsed jamming , continuous wave ( cw ) jamming and pseudo - noise ( pn ) jamming .
we also study the impact of jammer state information ( jsi ) , clipping and interleaving , under typical tc application constraints . in order to assess
how far the performance of the considered codes is from the theoretical limits , we extend the concept of shannon s sphere packing lower bound ( splb ) . through our analysis , we are able to identify which are the critical values of the signal - to - interference ratio ( sir ) for which the coding scheme is no more able to guarantee an acceptable level of protection . as further possible candidate schemes , we consider short parallel turbo codes ( ptc ) and extended bch ( ebch ) codes with soft - decision decoding . for decoding the ebch codes we consider the most reliable basis ( mrb ) algorithm , which has been successfully applied to these codes over the awgn channel .
we extend its use also to the jamming channel .
the organization of the paper is as follows . in section [ sec : two ] we introduce the types of jamming . in section [ sec : three ] we give examples of the performance of the bch@xmath2 code over the jamming channel . in section [ sec : four ] we describe the new coding schemes and in section [ sec : five ] the performance metrics adopted , including an extension of the splb . in section [ sec : six ] we provide some numerical examples . in section [ sec : seven ] we evaluate the impact of finite length interleavers for the system using ptc . finally , section [ sec : eight ] concludes the paper .
the definition of the types of jamming we consider is shortly reminded next .
we assume that the system adopts a direct sequence - spread spectrum ( ds - ss ) , and binary phase shift keying ( bspk ) modulation with carrier circular frequency @xmath10 .
the ds - ss is characterized by bandwidth @xmath11 and processing gain @xmath12 . moreover ,
the length ( period ) of the spreading sequence is denoted by @xmath13 .
a pulsed jamming signal has the following characteristics : * white gaussian noise on the whole bandwidth @xmath11 ; * discontinuity , with pulse active time @xmath14 and period @xmath15 , which means that the pulse is active for a fraction of time ( also called duty cycle ) @xmath16 ; * power @xmath17 during the active time @xmath14 , and zero for the remaining time @xmath18 . during the active time
the jamming signal has a power spectral density which is constant over the @xmath11 band with value @xmath19 , where @xmath20 . for proper comparison
it is also useful to introduce an equivalent ( with the same energy ) gaussian continuous jamming signal .
since the same energy is transmitted over @xmath15 instead of @xmath14 , it has a power @xmath21 .
this equivalent jamming signal has a power spectral density constant over the @xmath11 band , with value @xmath22 , where @xmath23 .
the error rate performance can be expressed in the terms of the ratio @xmath24 between the energy per bit and the equivalent one - side jamming spectral density .
when an error correcting code is applied , the impact of pulsed jamming can be mitigated through interleaving .
this is because most of the forward error correction schemes are designed for an awgn channel which exhibits no memory .
they do not handle bursts of errors .
an interleaver distributes a burst of errors among many consecutive codewords . by doing this
the number of errors contained in each codeword is limited , the code is able to correct them and the burst is neutralized . for analysis purposes
, we can initially refer to an ideal interleaver .
this implies that if a burst of errors corresponds to a fraction @xmath25 of the symbols , its impact after de - interleaving is modeled as a probability @xmath25 , for each symbol , of having a higher noise variance .
even if an ideal interleaver can not be implemented , it is very useful for analytically investigating the performance over jamming channels .
then , the performance in the presence of a real interleaver can be determined through simulation .
a cw jamming is a narrowband , continuous signal of type @xmath26 hence , it is a pure tone with : * circular frequency @xmath27 which , in general , may be different from the signal circular frequency @xmath10 ; * initial phase @xmath28 which , in general , may be different from the signal initial phase ( conventionally set to @xmath29 ) ; * power @xmath30 .
the worst case occurs when @xmath31 . under the hypothesis of having a large @xmath32 , the jamming contribution on a generic symbol
can be modeled by a gaussian random variable with zero mean and variance @xmath33 , where @xmath34 is the bit time duration . in the case of
cw jamming it is preferable to express the error rate performance in terms of the sir , @xmath35 , where @xmath36 is the signal power .
let us denote by @xmath37 the spreading sequence used in the ds - ss system .
a pn jamming is a signal of type @xmath38 hence , it is a ds - ss signal with : * circular frequency equal to @xmath10 ; * spreading sequence @xmath39 different from @xmath37 ( although it may have the same length ) ; * time delay @xmath40 on the spreading sequence ; * power @xmath30 .
a common choice for the spreading sequence @xmath37 is to use a gold code .
the novel cryptographic pn sequences proposed for tc applications are very long ( a suggested length is @xmath41 ) . once again , under the hypothesis of a large @xmath32 , the interfering contribution on a generic symbol can be modeled by a gaussian random variable with zero mean and variance @xmath42 .
thus we can obtain the same expression as for cw jamming with @xmath43 .
for all the types of jamming the channel is also impaired by thermal noise with signal - to - noise ratio per bit @xmath44 . as the two disturbances are independent one each other , when simultaneously present , their variances can be summed .
the tc protocols for synchronization and channel coding are specified ( with some differences ) both in the recommendation @xcite issued by the consultative committee for space data systems ( ccsds ) and in the standard @xcite issued by the european cooperation for space standardization ( ecss ) .
let us refer to the ccsds recommendation @xcite : it specifies the functions performed in the `` synchronization and channel coding sublayer '' in tc ground - to - space ( or space - to - space ) communication links . in short ,
the sublayer takes transfer frames ( tfs ) produced by the upper sublayer ( `` data link protocol sublayer '' ) , elaborates them and outputs communications link transmission units ( cltus ) that are passed to the lower layer ( ` physical layer' ) where they are mapped into the transmitted waveform by adopting a proper modulation format .
details on the structure of the tf and cltu can be found in @xcite , @xcite and are here omitted for the sake of brevity .
the current ccsds recommendation and ecss standard use a bch@xmath2 code for error protection against noise and interference . at the receiver
side a hard decision is taken on the received symbols .
the performance against pulsed jamming of the hard - decision decoded bch@xmath2 code is rather poor .
examples are shown in fig .
[ fig : standard_no_int ] for the case without the interleaver ( as addressed by the current standard ) and in fig .
[ fig : standard_with_int ] for the case with the ( ideal ) interleaver .
performance is expressed in terms of the cer ; the value of @xmath45 has been set equal to @xmath46 db and the cer is plotted as a function of @xmath47 for some values of @xmath25 . as expected , performance degrades for decreasing @xmath25 ; the use of the ( ideal ) interleaver introduces for @xmath48 an improvement that however remains unsatisfactory to the point that reaching cer = @xmath8 , that is a reference value for tc applications , is practically impossible for @xmath49 and @xmath50 .
code over pulsed jamming channel , for @xmath51 db and no interleaver .
[ fig : standard_no_int],width=340 ] code over pulsed jamming channel , for @xmath51 db and ideal interleaver .
[ fig : standard_with_int],width=340 ] an example of the performance of the hard - decision decoded bch@xmath2 code against cw jamming is shown in fig .
[ fig : standard_cw ] for @xmath52 , @xmath43 and @xmath53 ( worst case ) .
the sir value is assumed as a parameter and we see that a sir in the order of @xmath54 db makes the system unpractical .
a further reduction in the value of @xmath12 would require higher sir values ; for example , by assuming @xmath55 , the cer target of @xmath8 becomes practically unreachable just for sir @xmath56 db .
code over cw jamming channel ( @xmath43 and @xmath53 ) , for @xmath52 .
[ fig : standard_cw],width=340 ] for pn jamming it is possible to verify that under the gaussian approximation its impact is equivalent to that of cw jamming with @xmath43 and @xmath53 . in fig .
[ fig : standard_pn ] we have plotted the cer curve as a function of the sir for different values of @xmath12 and @xmath57 db . so this curve also applies to the worst case cw jamming with the same sir .
code over pn jamming channel , for fixed @xmath57 db . [ fig : standard_pn],width=340 ]
the codes proposed for tc channel coding updating have length greater ( @xmath58 and @xmath59 ) and rate smaller ( @xmath60 ) than the standard code .
also the code type has been changed , with the aim to introduce state - of - the - art codes .
the mostly addressed candidates , in this sense , are ldpc codes , both binary and non - binary .
recently , however , we have also shown that potential competitors can be ptc and even bch codes , if they are soft - decision decoded with maximum likelihood ( ml)-like algorithms characterized by limited complexity .
the main features of these schemes are reminded next .
a class of binary ldpc codes that is suitable for tc applications has been proposed by the national aeronautics and space administration ( nasa ) and is described in @xcite .
it is based on the adoption of three systematic short binary ldpc codes designed using protographs with circulant matrices .
soft - decision decoding can be realized by using the classic sum - product algorithm with log - likelihood ratios ( llr - spa ) .
non - binary ldpc codes with the same lengths have been analyzed by nasa @xcite and independently by the deutsches zentrum fr luft- und raumfahrt ( dlr ) in a joint work with the university of bologna ( unibo ) @xcite . in this paper
we refer to the dlr - unibo implementation .
decoding is realized by using iterative algorithms based on fast hadamard transforms .
parallel turbo codes are one of the coding options of the ccsds recommendation for telemetry ( tm ) links @xcite . the ccsds turbo encoder is based on the parallel concatenation of two equal 16-state systematic convolutional encoders with polynomial description @xmath61 .
the interleavers are based on an algorithmic rule proposed by berrou and described in @xcite .
the ccsds turbo encoder has four possible information frame lengths : @xmath62 and @xmath63 bits .
the nominal code rate can be @xmath64 and @xmath65 . however , higher rates are obtainable by puncturing @xcite . maintaining unchanged the encoder structure ,
we have considered frame lengths shorter than those in the tm recommendation and fixed the nominal code rate to @xmath60 , in such a way as to comply with the nasa s choices discussed above . because of the shorter length , we can not use the interleavers in @xcite and we must design new smaller interleaving structures . among a wide number of different options , we have focused attention on : completely random , spread @xcite , quadratic permutation polynomial ( qpp ) @xcite and dithered relative prime ( drp ) @xcite interleavers . moreover , since the constituent ccsds convolutional codes have @xmath66 states , four extra - tail bits are needed for termination ; then , the turbo codeword length is @xmath67 . as an example , for the case of @xmath4 , this implies to have @xmath68 and an actual code rate @xmath69 . in order to achieve exactly the code rate @xmath60 ,
as it is necessary for fair comparison , we have implemented a suitable puncturing strategy .
the interleaver and the puncturing pattern have been jointly optimized , in such a way as to maximize the minimum distance @xmath70 and minimize the codewords multiplicity @xmath71 ( i.e. , the number of codewords with hamming weight @xmath70 ) ; these parameters , in fact dominate the code performance at low error rates .
the results of the design optimization are shown in table [ tab : syst ] ; besides @xmath70 and @xmath71 , also the information multiplicity @xmath72 ( sum of the input weights over all the codewords with weight @xmath70 ) is provided since , together with @xmath70 , it determines the asymptotic bit error rate performance .
.selected interleavers for the parallel turbo codes .
[ cols="^,^,^ " , ] [ tab : syst ] the decoding of turbo codes is performed by iteratively applying the well - known bcjr algorithm @xcite to the constituent encoders .
although soft - decision decoding of bch codes is generally complex , in the case of short bch codes with high rate , an exact ml soft - decision decoding is possible , through its trellis representation ( for example , based on the viterbi or the bcjr algorithms ) .
however , having now decided to use codes with rate @xmath60 and the shortest length @xmath5 , these techniques are too involved and therefore can not be applied .
alternative solutions can be found , at least for the case of @xmath4 .
in fact , the ebch@xmath7 code can be efficiently decoded by using sub - optimal soft - decision decoding algorithms .
several options are available for this purpose , also exploiting ldpc - like code representations @xcite . for this code
, we have focused on the mrb algorithm @xcite , which has very good performance and acceptable complexity .
it consists of the following steps : 1 .
identify @xmath73 most reliable received bits and obtain from them a vector @xmath74 .
2 . construct a systematic generator matrix @xmath75 corresponding to these bits .
3 . encode @xmath74 by @xmath75 to obtain a candidate codeword @xmath76 .
4 . choose the order @xmath77 of the algorithm .
5 . consider all ( or a proper subset of ) test error patterns of length @xmath78 and weight @xmath79 .
6 . for each of them :
sum to @xmath74 , encode by @xmath75 , verify if the likelihood is higher than that of the previous candidate codeword and , if this is true , update the candidate .
further details can be found in @xcite and the references therein .
the mrb algorithm can be applied also to the other schemes ( to ldpc codes , in particular ) where , depending on the order @xmath77 , it can provide performance comparable to , or even better than , that offered by the iterative algorithms .
the performances of the new coding schemes presented in section [ sec : four ] have been recently discussed and compared over the awgn channel @xcite .
we have verified they can provide an advantage of more than @xmath80 db over the current bch@xmath2 code . in section [ sec : six ] we will show that similar improvements can be achieved against jamming . for soft - decision decoding , the knowledge of the jamming state can play a relevant role . for the case of pulsed jamming , for example , to have jsi means to know the noise variance for each symbol .
thus , in our simulations , we have considered the case of perfect jsi , which means the receiver is able to identify the fraction , @xmath25 , of symbols affected by jamming and the remaining fraction , @xmath81 , that is only affected by thermal noise , and properly estimate their noise variances , which are @xmath82 and @xmath83 , respectively . on the opposite side ,
we have also considered the case when jsi is not available , and the variance used for llr calculation of the decoder input is always equal to the average value @xmath84 .
the goodness of the proposed solutions can be measured through the distance of the cer curves from the splb . among the various approaches available to compute the splb
, the most suitable one is the so - called sp59 @xcite .
a modified version of this bound is also available ( called sp67 @xcite ) , that is able to take into account the constraint put by the signal constellation ( bpsk in the present analysis ) .
more recent improvements @xcite , @xcite are significant only for high code rates or long codeword lengths , and these conditions are not satisfied by the codes here of interest .
thus , in the present study , we consider the sp59 as the most significant splb .
the splb reported in the literature refers to the awgn channel and needs generalization .
this is easy to achieve for pulsed jamming on the condition that the pulse duration is a multiple of the codeword length and no interleaving is applied . for this purpose ,
let us denote by splb@xmath85 the sp59 bound for the awgn channel . under the hypotheses above an extended sphere packing lower bound , esplb , for the case of pulsed jamming
can be defined as : @xmath86 by setting @xmath87 in , we obtain an expression which is valid also for cw and pn jamming channels , when the gaussian approximation is applied . the esplb given by
will be considered in section [ sec : six ] as a useful benchmark for the case with jsi and without interleaving .
due to limited space , we focus on pulsed jamming and on codes with @xmath88 and @xmath4 . the analysis can be extended to the longer codes for which , however , the complexity issue for the decoding algorithms adopted can become more critical . the performances of the new coding schemes presented in section [ sec : four ] are compared in figs .
[ fig : results_noint_nojsi]-[fig : results_withint_withjsi ] assuming @xmath50 , @xmath57 db and variable @xmath45 , with and without an ( ideal ) interleaver , with and without jsi . in the latter case , to limit the impact of the incorrect noise estimation a clipping threshold , equal to twice the amplitude , has been applied to the signal at the channel output . for the ptc
we have used the optimal drp interleaver reported in table [ tab : syst ] .
the order used for the mrb algorithm applied to the ebch code is @xmath89 . without interleaving and without jsi .
[ fig : results_noint_nojsi],width=340 ] with interleaving and without jsi .
[ fig : results_withint_nojsi],width=340 ] without interleaving and with jsi .
[ fig : results_noint_withjsi],width=340 ] with interleaving and with jsi .
[ fig : results_withint_withjsi],width=340 ] for three of the considered scenarios the relative behavior of the proposed schemes is very similar : the best performance is achieved by the ebch code , while ptc and non - binary ldpc codes are very close one each other and suffer a penalty with respect to the ebch code that depends on the simulation conditions .
an interesting exception occurs for the case with interleaving and without jsi where , because of the ordering mechanism that is at the basis of mrb , the ebch code loses its leadership .
the ebch code is also very close to the esplb , where applicable ( see fig .
[ fig : results_noint_withjsi ] ) . on the contrary , the performance of the binary ldpc code is rather poor with a loss than can be larger than @xmath90 db with respect to the best solution .
these gaps are even more pronounced than those found over the awgn channel @xcite .
the results shown in the previous section for the systems using interleaving referred to the adoption of an ideal interleaver . in this section we discuss the effect of using real interleavers characterized by finite length .
we consider the particular case of using ptc , but a similar analysis could be developed for the other schemes . with reference to the short description given in section [ sec :
three ] ( but further details can be found in @xcite ) , we suppose that interleaving is applied at cltu level and also taking into account the possible presence of partitioning . this occurs when the tf has length @xmath91 that is not a multiple of @xmath78 : the tf is partitioned into @xmath92 input blocks and , if needed , zero filling is used to complete the last block .
each block is then encoded producing @xmath93 cltu coded bits . for the sake of simplicity , in this first evaluation we neglect the presence of the preamble ( @xmath66 bits ) and the postamble ( 64 bits ) that are added for cltu synchronization @xcite .
so , we assume that an interleaver is applied to the @xmath94 cltu coded bits to increase the protection against bursts ; it involves all the @xmath95 codewords of the cltu . as a simple example
, we consider a square @xmath96 row - by - column interleaver where @xmath97 ; the @xmath94 cltu coded bits are written by row and read by column ( or vice versa ) . in this case , wishing to compare the impact of bursts of growing length it is preferable to refer directly to the power @xmath17 and the corresponding ratio @xmath98 .
an example of the impact of a real interleaver on the transfer frame error rate ( fer ) in the considered scenario is shown in fig .
[ fig : real_int ] , for @xmath99 db , @xmath100 and a burst length of @xmath101 bits . we observe that the cltu interleaver is very effective against bursts produced by pulsed jamming . for long tfs ( like the one here considered )
the burst is practically neutralized by the interleaver . without interleaving and using a @xmath102 row - by - column interleaver , for a burst length of @xmath101 bits ; @xmath103 db and @xmath100 .
[ fig : real_int],width=340 ]
for the first time at our knowledge , the performance over jamming channels of new coding schemes that potentially replacing the bch@xmath2 code in the space tc channel coding standard has been investigated .
the adoption of soft - decision decoding , combined with interleaving and jsi , allows to achieve significant improvements .
whilst it is confirmed that non - binary ldpc codes , that are considered the most eligible candidates , are generally very good , we have shown that comparable performances can also be achieved by using ptc or ebch codes .
we have studied the case of @xmath88 and @xmath4 , where ebch codes often offer the best results at an acceptable complexity .
this fact , however , can not be used to draw general conclusions .
first of all , extending the sub - optimal decoding algorithms to longer codes , while maintaining acceptable complexity , may be quite difficult .
additionally , the sub - optimal algorithms generally define `` complete '' decoders .
as it is well known , this may be a penalty for the undetected codeword ( or frame ) error rate that , in tc applications , is at least as important as the codeword ( or frame ) error rate .
further work is progress to assess either the complexity and the completeness issues .
g. liva , e. paolini , t. de cola , and m. chiani , `` codes on high - order fields for the ccsds next generation uplink , '' in _ proc .
2012 6th advanced satellite multimedia systems conf .
( asms ) and 12th signal proc .
space commun .
workshop ( spsc ) _ , wessling , germany , sep .
l. costantini , b. matuz , g. liva , e. paolini , and m. chiani , `` non - binary protograph low - density parity - check codes for space communications , '' _ int .
commun . and networking _ , vol .
30 , pp . 4351 , 2012 .
g. p. calzolari , e. vassallo , f. chiaraluce , and r. garello , `` turbo code applications on telemetry and deep space communications , '' in _ turbo code applications : a journey from a paper to realization _ , k. sripimanwat , ed.1em plus 0.5em minus 0.4emspringer , 2005 , ch . 13 , pp .
321344 .
m. baldi , g. cancellieri , and f. chiaraluce , `` iterative soft - decision decoding of binary cyclic codes based on spread parity - check matrices , '' in _ proc .
softcom 2007 _ , split - dubrovnik , croatia , sep .
2007 .
m. baldi , m. bianchi , f. chiaraluce , r. garello , i. aguilar sanchez , and s. cioni , `` advanced channel coding for space mission telecommand links , '' in _ proc .
ieee 78th vehicular technology conference ( vtc fall 2013 ) _ , las vegas , nv , sep . 2013 . | the aim of this paper is to study the performance of some coding schemes recently proposed for updating the tc channel coding standard for space applications , in the presence of jamming . besides low - density parity - check codes , that appear as the most eligible candidates
, we also consider other solutions based on parallel turbo codes and extended bch codes .
we show that all these schemes offer very good performance , which approaches the theoretical limits achievable . error correcting codes , jamming , telecommands . |
galaxy morphology in different wave - bands provides useful information on the nature of galaxy evolution as well as the overall distribution of galaxy constituents such as old red giants , young luminous stars , gas , dust etc .
for example , the younger population i stars associated with massive gas - rich star formation regions light up the disk galaxies in optical wavelengths .
the distribution of older population ii stars , the dominant matter component near the central regions of galaxies , remains hidden .
the presence of interstellar dust hides the old stellar population especially in late - type disk galaxies .
the nir light , on the other hand , is much less affected by the interstellar dust and more sensitive to the older populations .
thus it provides a penetrating view of the core regions in disk galaxies . therefore careful analysis of morphological differences between the optical and infrared images would not only provide valuable insight into the role of population classes in morphology but also reveal whether the discrepancies are due to singular or combined effects of extinction and population differences ( jarrett et al .
2003 ) . in the morphological studies of cosmological objects
the most widely used technique is the ellipse - fitting method , ( carter 1978 ; williams & schwarzschild 1979 ; leach 1981 ; lauer 1985 ; jedrzejewski 1987 ; fasano & bonoli 1989 ; franx , illingworth & heckman 1989 ; peletier et al .
1990 ) . in this study
we use a set of measures known as the minkowski functionals ( hereafter mfs , minkowski 1903 ) to analyze the morphology of nir galaxies .
contrary to the conventional method , the mfs provide a non - parametric description of the images implying that no prior assumptions are made about the shapes of the images .
the analyses based on the mfs appear to be robust and numerically efficient when applied to various cosmological studies , e. g. , galaxies , galaxy - clusters , cmb maps etc .
( mecke , buchert & wagner 1994 ; schmalzing & buchert 1997 ; kerscher et al .
1997 ; schmalzing & gorsky 1998 ; hobson , jones & lasenby 1999 ; novikov , feldman & shandarin 1999 ; schmalzing et al .
1999 ; beisbart 2000 ; kerscher et al . 2000a ; novikov , schmalzing , & mukhanov 2000 ; beisbart , buchert & wagner 2001a ; beisbart , valdarnini & buchert 2001b ; kerscher et al .
2001b ; shandarin 2002 ; shandarin et al .
2002 ; sheth et al . 2003 ; rahman & shandarin 2003 , hereafter paper 1 ) this is the second in a series of papers aimed to study the morphology of galaxy images using a set of measures derived from the mfs . in this paper
we analyze a larger sample of 2mass galaxies imaged at @xmath7 and @xmath3 band in nir ( jarrett 2000 ; jarrett et al .
2000 ; jarrett et al . 2003 ) .
we have described and tested the set of minkowski parameters derived from the two - dimensional scalar , vector and several tensor mfs to quantify galaxy shapes for a small sample of 2mass images in paper i. the analyses in paper i used contour smoothing to reduce the effect of background noise .
we have used the same technique in the present sample which contains nir galaxies over the entire range of hubble types including ellipticals , lenticulars and spirals .
the present investigation is aimed at obtaining structural information on 2mass galaxies by measuring their shapes quantified by ellipticity and orientation . as dusty regions of galaxies
become transparent in the nir , the imaging in this part of the spectrum should provide a clear view of the central core / bulge regions of these objects .
a systematic study of nir images should provide valuable information regarding the central structures of galaxies ( e. g. , optically hidden bar ) which would otherwise remain absent when viewed in optical wavelengths .
if only the old red giants illuminate galaxies at nir wavelengths and are decoupled from population i star lights , then the nir galaxies should show weak isophotal twist in their orientations compared to those in the visual wavelengths
. therefore it would be interesting to check whether or not isophotal twist is a wavelength dependent effect .
the organization of the paper is as follows : the 2mass sample and selection criteria are described in @xmath8 , a brief discussion of the parameters is given in @xmath9 .
we discuss the robustness of the measures to identify and discern galaxy isophotes of various shapes and present our results @xmath10 .
we summarize our conclusions in @xmath11 . in the appendix ( @xmath12 )
we demonstrate the sensitivity of several minkowski measures to image contamination by foreground stars .
the 2mass catalogue contains near - infrared images of nearby galaxies within redshift range from @xmath13 km s@xmath14 to @xmath15 km s@xmath14 .
the survey utilizes the nir band windows of @xmath16 , @xmath17 and @xmath18 ) .
the 2mass images have 1@xmath19 pixel resolution and @xmath20 beam resolution .
the seeing fwhm values for these images are typically between @xmath21 and @xmath22 in all three bands . for details of the 2mass observations , data reduction and analysis ,
readers are referred to jarrett ( 2000 ) and jarrett et al .
( 2000 , 2003 ) .
our sample contains 112 galaxies imaged in nir @xmath1 , @xmath2 , and @xmath3 bands .
it includes 32 elliptical , 16 lenticulars , and 64 spirals acquired from the 2mass extended source catalogue ( xsc ; jarrett et al .
the spiral sample contains 19 normal ( sa ) , 21 transitional ( sab ) , and 24 barred ( sb ) galaxies .
the galaxy types are taken from the rc3 catalogue ( de vaucouleurs et al .
the sensitivity and resolution ( @xmath23 ) of the nir data obtained in the 2mass is not adequate to derive independent galaxy sub - classification ( jarrett 2000 ) , therefore , we rely on the morphological classification based upon optical data derived in combination with both imaging and spectroscopy .
we construct the sample by hand with a moderately large number of galaxies of each type to make statistical inferences .
the primary motivation behind constructing the sample is to make a comparative analysis with previous results and to investigate the galaxies with new tools to gain further insight into overall galaxy morphology in infrared wave - bands .
we consider only bright galaxies in three different bands ; the @xmath3 band total magnitude for the galaxies is @xmath24 .
spiral galaxies with inclination up to @xmath25 are included in the sample .
no deprojection has been made to any of these galaxies prior to the analysis since the projection effect does not pose a serious threat to the reliability of the analysis when using a parameter such as ellipticity in the structural analysis of low inclination spiral galaxies ( martin 1995 ; abraham et al .
1999 ) .
all galaxies in our sample are flat fielded and background subtracted . except for three ellipticals , foreground stars have been removed from the rest of the sample .
those galaxies where a foreground star is left embedded in images are included purposely to illustrate the sensitivity of the morphological measures , as explained in the appendix .
for an object with arbitrary shape a complete morphological description requires both topological and geometrical characteristics . the mfs consist of a set of measures carrying both geometric ( e. g. , area , perimeter ) and topological ( the euler characteristic , ec ) information about an object .
the functionals obey a set of properties such as motion invariance , additivity and continuity ( see schmalzing 1999 ; beisbart 2000 ) .
for this study we derive morphological parameters using a selection of two - dimensional scalar , vector , and tensor mfs as described in paper i. we treat every image as a set of contour lines corresponding to a set of surface brightness levels .
a contour is constructed by linear interpolation at a given level . for every contour ,
the first step of the functional analysis provides three scalars : @xmath26 , @xmath27 , and ec ( also represented by the symbol , @xmath28 ) ; three vectors or centroids : @xmath29 , @xmath30 , and @xmath31 ; and a total of nine components of three symmetric tensors @xmath32 , @xmath33 , and @xmath34 . here
@xmath35 ( for details see paper 1 ) . in the next step ,
the eigenvalues ( @xmath36 and @xmath37 ; @xmath38 ) of the tensors are found taking centroids as the origins of corresponding tensors . after calculating the eigenvalues
, we proceed to construct the axes and orientations of the `` auxiliary ellipse '' ( hereafter ae ) . to construct the area tensor ae , for example , we take the eigenvalues of @xmath32 and ask what possible ellipse may have exactly the same tensor . when we find that particular ellipse , we label it as the area tensor ae .
the orientation of the semi - major axis of the ae with respect to the positive x - axis is taken as its orientation .
the aes corresponding to the perimeter and ec tensors are constructed in a similar manner . to discern morphologically different objects , therefore ,
we use ellipticities ( @xmath39 ) and orientations ( @xmath40 ) of the aes rather than the eigenvalues of the tensors .
we define ellipticity of the aes as @xmath41 where @xmath42 corresponds to one of the three tensors , and @xmath43 and @xmath44 are the semi - major and semi - minor axes of the aes .
the use of aes effectively relates a contour to an ellipse : the similarity of three aes is a strong evidence that the shape of the contour is elliptical .
for example , in case of a perfect elliptic contour , the aes will be the same .
in particular , the areas of all three aes will be equal to the area of the contour , i. e. , @xmath45 , and the perimeters of the ellipses will be equal to the perimeter of the contour , i. e. , @xmath46 .
in addition , the orientations of all three ellipses will coincide with the orientation of the contour .
therefore , if plotted , all three aes will be on top of each other , overlapping with the contour .
for that contour , all three vector centroids will also coincide with each other and with the center of the contour .
note that the latter alone does not guarantee that the contour itself is elliptical in nature since for any centrally symmetric contour the centroids would coincide .
however , for a non - elliptical contour all three aes will be different in size and orientation ( see fig . 1 , paper i ) .
note that the sets of eigenvalues from three tensors can be used to construct three `` anisotropy '' parameters instead of three aes ( see also beisbart 2000 ) .
the parameters can be defined as @xmath47 where @xmath42 has the same meaning as before . to better understand the behaviors of @xmath39 and @xmath48 , we show these parameters in fig . [ scaling ] as functions of contour area ( @xmath26 ) for four elliptic profiles of different flattening .
panel 1 of the figure shows ellipticity for all four profiles .
panels 2 , 3 , and 4 show , respectively , @xmath49 , @xmath50 , and @xmath51 where each of these panels also has four profiles . for each profile , we find that the @xmath39 of the aes are identical and coincide with each other ( panel 1 ) .
the @xmath48 , on the other hand , do not coincide even for perfect elliptic contours ( panels 2 , 3 , and 4 ) .
we also find that the relative separations between different @xmath48 change with the flattening of the contours . from the behavior of the parameters one can think that the @xmath39 of the aes act as parameters that are scaled with respect to the @xmath48 . for contours with arbitrary flattening
, the @xmath48 from the area and ec tensors need to scale down to match with the @xmath39 of the respective aes .
the @xmath52 from the perimeter tensor , however , needs to scale up for spherical and moderately elongated contours . for highly elongated contours , however , all three @xmath52s need to scale down .
the illustration of the `` anisotropy '' parameter serves two purposes .
first , it gives us a feeling of the aes compared to the conventional parameter that deals with eigenvalues .
second , it demonstrates that one can derive various shape measures from the set of mfs .
apart from this parameter , one can also derive the shapefinder statistic as suggested by sahni , sathyaprakash & shandarin ( 1998 ) .
however , we restrict ourselves to ellipticity and orientation since these are the two widely used measures in astronomy
. we will explore the sensitivity and robustness of @xmath48 in our future work on optical galaxies .
the non - parametric approach for shape analysis , such as moments technique , has been known to the astronomical community for some time ( carter 1978 ; carter & metcalf 1980 ) .
it should be mentioned here that the morphological analyses based only on the moments of inertia would provide incomplete and sometimes misleading results . as an example , let us assume that one has a galaxy image which has been kept in a black box and analyzed using simply the inertia tensor without having a priori knowledge of the shape of the image .
the analysis based only on the moments of inertia will provide a resultant ellipticity of the object regardless of its actual shape . using this result
one can always infer an elliptical shape for the unseen object .
if one raises the question of the likeliness of the elliptical shape of the object , the analysis based on the inertia tensor alone will not be able to give a satisfactory answer .
one needs to invoke additional measure(s ) in order to justify the result .
it is at this point where the measures derived from the set of mfs appear to be effective .
subsequent analyses of the image using moments of the perimeter and ec tensor enables one to pin down the type of the galaxy and thus ensures the objectivity of the analysis .
the ellipticities obtained from different aes provide information ( regarding shapes ) similar to the conventional shape measure based on inertia tensor .
the main difference is that the conventional method finds the eigenvalue of the inertia tensor for an annular region enclosing mass density or surface brightness .
the method based on mfs , however , finds the eigenvalues of contour(s ) where the region enclosed by the contour(s ) is assumed to be homogeneous and to have constant surface density . in order to reduce the effect of noise present in the image we use a simple smoothing technique . instead of smoothing the whole image , we smooth contours at each brightness level using the procedure known as the unequally weighted moving average method .
the goal of this smoothing is to restore the initial unperturbed contour as much as possible and measure its morphological properties .
the implementation of the smoothing is described in detail in paper i. contour smoothing considerably improves the estimates of ellipticity derived from the tensor functionals ( see figs . 8 and 9 ,
paper i ) , therefore , in this study , we focus on the results obtained only from the smoothed contours .
however , we note two effects that arise as a result of smoothing
. first , in the outer regions of galaxies a smoothed , outer contour crosses the inner one .
it happens occasionally ; however , the area within the smoothed , outer contour remains greater than the inner one and does not affect the measurement .
second , in case of very large contour , we loose information .
the smoothing technique is an iterative process depending on the number of points along a contour ( see paper 1 for details ) .
a highly irregular contour that consists of large number of points eventually shrinks to a point due to excessive smoothing . as a result
we do not get any contribution from it .
however , this does not pose any threat on getting relevant information of the shape of image contours since we can always compare with the unsmoothed profiles to see how much information is lost .
in this paper we use a different notation to express ellipticity than in paper i where we used @xmath53 in the range 0 to 10 . the symbol @xmath54 for the ellipticity parameter is often used for
the definition of the hubble types for ellipticals and this symbol and range can be misleading when used as a parameter for other types of galaxies ( e. g. , spirals ) .
we , therefore , use @xmath39 as the characteristic of shape with the range 0 to 1 . the ellipticity and orientation can be used to isolate galaxies into two broad bins : non - barred and barred systems . for finer distinction
we need to use other parameters such as luminosity , color , surface brightness , half - light radius , etc .
both of these parameters have been applied to quantify properties of elliptical galaxies ( e.g. carter 1978 ; williams & schwarzschild 1979 ; leach 1981 , lauer 1985 ; jedrzejewski 1987 ; fasano & bonoli 1989 ; franx , illingworth & heckman 1989 ; peletier et al .
in many recent works this set of parameters has been used to obtain information on different structural components , e. g. , bar , bulge , disk etc . of disk galaxies ( athanassoula et al .
1990 ; martin 1995 ; wozniak et al .
1995 ; rozas , knapen & beckman 1998 ; abraham & merrifield 2000 ; laurikainen , salo & rautainen 2002 ; peng et al .
2002 , erwin & sparke 1999 ; erwin & sparke 2003 ; michel - dansac & wozniak 2004 ) .
galaxies with different morphologies appear to show characteristic signatures in the ellipticity profile .
for example , the profile of elliptical galaxies is generally monotonic .
a barred galaxy , however , shows a distinct peak with a spherical central region , signaling the presence of a central bulge and a bar . on the other hand , for a multi - barred system
several peaks appear in the profile ( wozniak et al . 1995 ; erwin & sparke 1999 ) .
elliptical galaxies usually show twists ( i. e. , a change in the orientation of ae ) in the orientation profiles with varied strength depending on the flattening of the contours .
barred galaxies , on the other hand , not only have large twists in their orientations but also have characteristic features such as a sharp peak , or two different but approximately constant orientations with an abrupt change in between ( wozniak et al . 1995 ; erwin & sparke 1999 ) . in the contemporary studies , therefore , a barred galaxy is identified as the system whose ellipticity and orientation profile simultaneously show the distinctive signatures mention above . in this study
we follow the same criterion to analyze disk galaxies , i. e. , we identify bars by visual inspection with the condition that both ellipticity and orientation profiles show the characteristic features simultaneously .
however , identifying disk galaxies as barred systems with the above criterion ( as used in previous studies mostly in optical wavelengths ) should be used with caution .
our experience shows that optically classified barred galaxies appear with distinct peaks in their ellipticity profiles but show continuous orientation over the region where the peak in the shape profiles persists .
therefore we feel that more elaborate treatment is needed for identifying barred systems rather than simply relying on the behaviors of ellipticity and orientation . in this study , therefore , when we encounter these types of systems , we refrain from drawing any conclusions about these galaxies .
we will explore the detailed shape properties of these galaxies in our future study including other structural measures such as fourier decomposition technique ( quillen , frogel & gonzalez 1994 ; buta & block 2001 ) .
we measure ellipticity and orientation for a set of contours obtained at different surface brightness levels .
generally these two parameters vary with the area of the contour , i. e. , @xmath55 and @xmath56 .
we analyze each galaxy at 30 different brightness levels where the levels correspond to equally spaced areas on a logarithmic scale covering almost the entire region of each galaxy . at every brightness level , contours are found and subsequently smoothed .
all three aes are constructed from the smoothed contours . in order to reduce information content
we present our final results using only the area tensor ae and hence drop the subscript @xmath42 .
below we briefly discuss the justification of the choice . for each galaxy , therefore , we show @xmath4 and @xmath5 of this particular ae as a function of contour area ( @xmath26 ) . for each galaxy ,
a thick dashed line is used to show the mean values of @xmath4 and @xmath5 calculated from @xmath7 and @xmath3 bands .
the mean values are used to estimate the overall change ( @xmath57 and @xmath58 ) which is defined as the difference between the highest and lowest value of the corresponding mean value .
it is a single number and independent of @xmath26 .
two thin solid lines are used to show the maximum and minimum of the measures obtained from different bands .
the difference between these two thin solid lines is used to quantify the scatter ( @xmath59 and @xmath60 ) in the parameters .
the @xmath61 varies at different regions of a galaxy and so depends on @xmath26 .
we use it as an indicator to measure the dependence of galaxy shapes on different colors . for each sample of galaxies ,
the result is presented in increasing order of 2mass ellipticity obtained from the 2mass catalogue .
the 2mass ellipticity and orientation are measured for the @xmath62 isophote in the @xmath3 band ( @xmath63 is the rms amplitude of the background noise provided by the catalogue ) . for each galaxy , the 2mass shape parameter is , therefore , a single number .
our analysis , on the other hand , provides a range of values obtained at different regions ( recall that @xmath4 or @xmath5 is a function of area ) from the galactic center .
we use the 2mass estimate of ellipticity and position angle as the reference .
it is shown by the horizontal dashed line . in all plots the vertical dashed line represents the contour area ( @xmath26 ) corresponding to @xmath3 band @xmath62 isophote .
we rescale the orientation profiles of a few galaxies to fit the desired ranges that have been chosen to show the @xmath5 profiles .
note that our results presented below correspond to the area @xmath64 . the range is chosen in order to exclude discreteness effects due to the grid .
the deviation ( @xmath65 ) as well as the scatter ( @xmath61 ) in the parameters , therefore , will be considered in this range . from section @xmath66 we know that for a perfect elliptic contour , the areas enclosed by the aes will be the same whereas for non - elliptic contours they will be different
. therefore , comparing the areas enclosed by different ellipses one can probe to which extent the shape of a contour can be approximated by an ellipse .
galaxies appear more regular in nir wave bands than in the visual wavelengths .
for example , a galaxy may appear grand in the visual bands with its giant spiral arms .
however , it will lose much of its grandeur in infrared wave - bands .
since spiral arms consist mostly of gas and young bright stars , they will be absent in long wavelength parts of the spectrum . to see up to
what extent galaxy contours of different hubble types retain their characteristic signatures in nir and how tensor ellipses help us to understand their shapes , we draw readers attention to figs .
[ aa_1 ] and [ aa_2 ] . in these figures
we show the relative difference in areas enclosed by three different aes as a function of contour area ( @xmath26 ) for a selection of elliptical and spiral galaxies .
the figures highlight only the interesting parts along the vertical axis .
each panel contains a total of nine curves : three curves corresponding to three aes from each band .
the dark , medium and light colors represent , respectively , the @xmath1 , @xmath2 , and @xmath3 band .
we show the area , perimeter , and ec ellipses , respectively , by the solid , dotted , and dashed - dotted lines . apart from three galaxies that are marked by `` s '' , the isophotal contours of most of the galaxies in fig .
[ aa_1 ] show elliptic nature in the nir .
few of these , e. g. , ngc 3158 and ngc 2778 ( galaxy number 8 and 15 respectively ) appear slightly non - elliptic in one or two bands .
the galaxies marked by `` s '' ( number 1 , 4 , and 7 ) are galaxies which have foreground stars embedded in their image . for these galaxies
the notable feature is the sharp increase in area around the region of the images where the star is embedded .
one can see that the same ellipse overlaps in different bands .
this is most notable from the ec ellipse ( dashed - dotted line ) since the ec tensor is the most sensitive to any disturbance along the contour .
the other two ellipses also behave in a similar manner but with less sensitivity .
this particular behavior shows that the aes detect unusual features attached to otherwise perfectly elliptical image body .
for example , when we compare ngc 3158 or ngc 2778 with those three galaxies , we notice that @xmath67 not only differs in different bands but it also spreads out differently around the edge .
it tells us that the contours of these galaxies are simply non - elliptic around the edge without any abnormal feature . when we compare galaxies of different hubble types , we find that almost all spiral galaxies ( fig . [ aa_2 ] ) show more non - elliptic nature than the elliptical galaxies
. the non - elliptic nature of spiral galaxy isophotes is reflected strongly in the behavior of the ec ellipse in all bands . for these galaxies ,
the ec ellipse is different not only from other two ellipses but it is also spread out arbitrarily in different bands . from figs .
[ aa_1 ] and [ aa_2 ] , it is clear that all three aes are useful for better characterization of a galaxy image .
the information provided by the aes definitely help to get finer distinction of galaxies . since our current interest is to focus on the gross morphological features rather than looking at the finer details of galaxy isophotes , to reduce the information content , therefore , we will only use the area tensor ae to present our final results .
the sample of elliptical galaxies has been divided into two groups .
group 1 contains galaxies with small ellipticities ( @xmath68 ) while group 2 has galaxies with @xmath6 . to simply locate a galaxy within a group , we label them with an integer number .
for example , ngc 4374(2 ) means that ngc 4374 is the second galaxy in its relevant group . in each group galaxy number 1
has the lowest 2mass ellipticity whereas number 16 has the highest value .
we follow the same format for all types of galaxies .
note that below we highlight only those galaxies that have interesting / unusual morphological features .
the general trends of galaxies will be investigated in @xmath11 .
* group 1 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellpe1 ] .
ngc 4374(2 ) and ngc 4261(9 ) , have uncommon profiles compared to other galaxies .
these galaxies are more elongated in the region @xmath69 than either around the center or near the edge which makes their profiles look `` centrally arched '' ( marked by `` a '' ) .
this particular type of behavior is shown only by sphericals and is absent in elongated galaxies .
the following three galaxies are shown with an `` s '' mark in fig .
[ ellpe1 ] : ngc 4278(1 ) , ngc 3193(4 ) , and ngc 3379(7 ) .
these galaxies have sharp kinks in the @xmath4 profiles .
these sharp changes are caused by a foreground star and do not correspond to any structural feature ( see appendix ) .
the rest of the galaxies show profiles typical of ellipticals .
galaxies in this group , in general , have small scatter . * orientation : * the orientation profiles of galaxies are presented in fig .
[ ange1 ] .
ngc 4458(3 ) is stable from the center up to @xmath70 and shows a large deviation near the edge .
ngc 3193(4 ) shows a notable peak in its orientation .
its profile shows a dip around the region where the foreground star is embedded .
ngc 3379(7 ) does not show any unusual feature in its profile .
its position angle , as well as the scatter , gradually increases towards the galactic center .
ngc 3379(7 ) has larger scatter at all distances from the galactic center than ngc 4278(1 ) and ngc 3193(4 ) .
note that the presence of a star on the galaxy contours may or may not be detected by the @xmath5 profile . the @xmath4 profile , on the other hand , is sensitive to any kind of disturbance on the contour .
if the unusual feature on the contour attributed by the star happens to be along the major axis , which is the case for ngc 3379(7 ) , it remains unnoticed and no peak appears in the @xmath5 profile .
this is the reason why the presence of a star on the contours may be missed as a signature on the @xmath5 profile . on the other hand ,
if the feature is slightly off the major axis , it changes the orientation of the contour significantly with respect to the inner and outer contours .
this is the case for ngc 4278(1 ) and ngc 3193(4 ) . in spite of being a spherical galaxy ( e1 + ) , ngc 4494(6 )
has remarkably low twist ( @xmath71 ) and scatter in its orientation .
it is very unusual because a spherical contour is highly susceptible to background noise .
a slight perturbation caused by the noise distorts the contour s shape and the direction of perturbation becomes its orientation .
vcc 1737(10 ) , ngc 3159(12 ) , and ngc 3226(13 ) have orientation profiles quite similar to disk galaxies as shown later in this section .
the @xmath5 profiles of vcc 1737(10 ) and ngc 3226(13 ) appear `` arch like '' where the latter galaxy has a longer arch than the former .
these three galaxies also have large scatter in their orientations .
* group 2 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellpe2 ] . galaxies with @xmath6
show a `` centrally spherical '' trend in their profiles .
the variation in the profile becomes stronger with the increasing flattening of the galaxies .
the overall change in @xmath4 observed from the center to the edge varies from galaxy to galaxy .
it changes from as low as @xmath72 ( ngc 315(2 ) ) to as high as @xmath73 ( ngc 5791(15 ) ) .
the measurements for most of the galaxies in all three bands nicely coincide . as a result
the galaxies in this group show the lowest amount of scatter .
* orientation : * the orientation profiles of galaxies in this group are shown in fig .
[ ange2 ] .
the flattened ellipticals have quite stable orientations with relatively low twist and scatter compared to those in group 1 . for all galaxies
the @xmath58 is within @xmath74 .
the relatively small twist observed in the elliptical galaxies with @xmath6 suggests that flattening and isophotal twists are most likely anti - correlated in the nir wave - bands ( see @xmath12 and fig .
[ galletta ] for more ) .
this result is in agreement with the ( optical ) correlation found by galletta ( 1980 ) who noted that the maximum apparent flattening and the highest observed twist are inversely related .
the sample of lenticulars contains 16 galaxies .
the ellipticity and orientation of these galaxies are shown in figs .
[ ellps0 ] and [ angs0 ] respectively .
* ellipticity : * for ngc 1315(1 ) , our estimate indicates a spherical inner part ( @xmath75 ) in between @xmath76 , suggesting a spherical bulge for the galaxy .
ngc 3598(7 ) , ic 4064(8 ) , ngc 3816(12 ) , ngc 3768(13 ) , and ngc 2577(16 ) have profiles quite similar to elliptical galaxies : a smoothed variation in elongation with radius .
* orientation : * the orientations of lenticular galaxies generally have larger scatter than the elliptical galaxies .
ngc 4659(4 ) shows the largest twist in this sample , ( @xmath77 ) ; its orientation experiences a large change within @xmath78 .
lenticular galaxies with ( 2mass ) ellipticity @xmath79 ( from galaxy number 8 and above in fig .
[ angs0 ] ) are observed to have less scatter than their spherical counterparts .
this property is quite similar to flattened ellipticals .
the scatter is large for lenticulars that are spherical in shape and it varies at different regions from the galaxy centers .
@xmath80 both shape and orientation profiles of ngc 4620(9 ) and ngc 2544(14 ) exhibit a bar like feature .
these galaxies are labeled as `` ib '' where `` ib '' stands for `` infrared bar '' .
the sample of spirals has been divided into four groups based on the degree of scatter ( @xmath59 ) observed in the @xmath4 profiles .
@xmath59 is measured for all @xmath26 in the range @xmath64 .
group 1 contains spirals that have the least scatter while group 4 has galaxies with the largest scatter .
galaxies with @xmath81 are included into group 1 , for group 2 the range is @xmath82 , those with @xmath83 are in group 3 , and finally galaxies with @xmath84 define group 4 .
the grouping of the spiral sample has been done by visual examination of the ellipticity profiles and therefore it is quite crude ; our intention is simply to highlight the interesting features apparent in the shape profiles of spiral galaxies viewed in the nir . in these groups ,
galaxies are organized by increasing order of ( 2mass ) ellipticity .
* group 1 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellps1 ] .
only four galaxies are optically classified ( rc3 ) as barred systems .
these are ngc 4262(2 ) , ngc 4024(3 ) , ngc 3384(10 ) , and ngc 4394(13 ) . except the last one ,
the shape profiles of other three galaxies posses distinctive feature , i. e. , a clear appearance of peak(s ) in the profiles , indicating ( qualitatively ) similar morphology both in optical and nir .
these three galaxies are labeled as `` oib '' where `` oib '' stands for `` optical and infrared bar '' .
we label ngc 4394(13 ) as `` ob '' where ` ob' stands for `` optical bar '' only .
we give our explanation below for this labeling .
ngc 4151(6 ) is a highly spherical galaxy compared to other members of this group .
it has @xmath85 from the galactic center up to @xmath86 ; beyond this limit the ellipticity increases a little bit .
this galaxy , an intermediate ringed transitional spiral , is most likely dominated by a very large spherical bulge .
the profiles of ic 3831(8 ) , ngc 4143(9 ) , ngc 3418(11 ) , and ngc 5326(12 ) are quite similar to elliptical galaxies .
ngc 3912(15 ) and ugc 5739(16 ) are two highly flattened galaxies in this group .
their profiles appear `` centrally arched '' , an unusual feature shown by several nir elliptical galaxies .
* orientation : * the orientation profiles of galaxies in this group are presented in fig .
[ angs1 ] .
the overall twists ( @xmath58 ) in these galaxies range in between @xmath87 .
ngc 4151(6 ) , a galaxy of spherical shape , shows the largest scatter in its orientation .
since highly spherical contours can be perturbed easily by noise , it is not surprising that the galaxy would have large variations in its orientation at different wave - bands .
the profile of ngc 4143(9 ) is quite similar to the barred galaxy ngc 3384(10 ) , although its orientation changes smoothly compared to that of ngc 3384(10 ) .
@xmath80 the profiles of ngc 3177(4 ) and ngc 3684(5 ) show a bar like feature although these galaxies are optically classified as non - barred systems .
however , we label these galaxies as `` ib '' after careful visual inspection of their unsmoothed and smoothed profiles . the profiles of ngc 4651(7 ) and ngc 4494(13 ) deserve attention because of their complex morphologies .
the ellipticity profile of ngc 4651(7 ) has a peak that is confined within @xmath88 . in this region
its orientation is different from the bulge and the disk .
since it is an early type spiral with @xmath89 subclass , we think the feature is due to the ring . according to rc3 classification ngc 4494(13 )
is a barred galaxy .
but from the nir profile it is not clear whether the galaxy really possesses a bar like structure .
the galaxy is devoid of any distinct peak in its shape profile even - though it has considerable amount of twist ( @xmath90 ) .
it shows higher flattening near the edge which is probably the elongation of the disk .
we are uncertain about the actual shape and therefore label the galaxy as `` ob '' ( optical bar ) .
* group 2 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellps2 ] .
it contains seven galaxies that are optically classified as barred systems .
the galaxies are ngc 3974(1 ) , ic 357(7 ) , ngc 4662(10 ) , ngc 6347(11 ) , ngc 5737(13 ) , ugc 2585(14 ) , and ic 568(16 ) .
these galaxies show a bar like feature in their nir ellipticity profiles .
we , therefore , label them as `` oib '' .
the profiles of ngc 3277(2 ) and ngc 4714(8 ) are quite similar to elliptical galaxies .
* orientation : * the orientation profiles of the galaxies in this group are shown in fig .
[ angs2 ] .
the overall twists range between @xmath91 .
the profiles of ngc 3974(1 ) and ic 357(7 ) have similar characteristics : a constant direction along the central bars while the orientation changes in the central bulge and outside disk . @xmath80 the following three galaxies are optically - classified non - barred spirals : ngc 4146(3 ) , ic 863(6 ) , and ugc 5903(9 ) . however , these galaxies show bar like features in their nir profiles and therefore we label them as `` ib '' .
note that the orientation profile of ngc 4152(4 ) has two distinct peaks .
its ellipticity profile , however , does not possesses any feature to be considered as a barred galaxy .
therefore we are not certain whether or not it can be considered as a barred galaxy .
the orientation profile of ngc 5478(12 ) shows two different directions .
its shape changes considerably around the region where the position angle changes .
the variation in ellipticity is extended over a wide region . from visual inspection
a question can be raised about whether or not the galaxy is a barred system .
a careful inspection of both unsmoothed and smoothed profiles indicates that it is not a barred galaxy .
* group 3 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellps3 ] .
this group contains the largest number of optically classified barred galaxies : ngc 5716(3 ) , ugc 3936(4 ) , ngc 4375(6 ) , ugc 7073(9 ) , ngc 3811(11 ) , ic 742(12 ) , ngc 5260(13 ) , ic 1764(15 ) , and ugc 3839(16 ) . except for ngc 4375(6 ) and ugc 7073(9 ) , the nir profiles of other 7 galaxies show similarity ( at least qualitatively ) with their optical morphology ( `` oib '' ) .
the profiles of both ngc 4375(6 ) and ugc 7073(9 ) lack characteristic signatures in their profiles and therefore have been labeled as `` ob '' .
* orientation : * the orientations of galaxies are shown in fig .
[ angs3 ] .
the overall twists range between @xmath92 .
the twist as well as scatter in the orientations of these galaxies are high compared to those in the previous two groups
. among `` oib '' galaxies , only ugc 3936(4 ) and ugc 3839(16 ) have large twists in their orientations . for other `` oib '' galaxies ,
the twists are relatively small .
@xmath80 the following three galaxies are optically classified as non - barred spirals : ugc 2303(1 ) , ugc 3053(2 ) , and ngc 3618(8 ) .
the profiles of these galaxies in nir , however , show features similar to barred systems .
we label these galaxies as `` ib '' after visual inspection of their unsmoothed and smoothed profiles .
* group 4 * * ellipticity : * the ellipticity profiles of galaxies in this group are shown in fig .
[ ellps4 ] . according to rc3 catalogue ,
the following four galaxies are barred systems : ugc 2705(4 ) , ugc 3171(9 ) , ngc 5510(13 ) , and ugc 1478(16 ) .
it is interesting to note that the nir profiles of none of these galaxies have characteristic features resembling bar like systems .
therefore we label these galaxies as `` ob '' .
* orientation : * the orientations of galaxies in spiral group 4 are shown in fig .
[ angs4 ] .
the overall twists range between @xmath93 .
galaxies in this group have the highest scatter in their orientations compared to those in the other three groups . among `` ob '' galaxies , ugc 1478(16 )
has a unique profile .
the galaxy has a distinct , highly flattened , central bar .
its orientation is constant along the central bar and changes sharply to the orientation of the disk .
@xmath80 the following three galaxies are optically - classified non - barred systems : ngc 1219(1 ) , ugc 3091(8 ) , and ugc 7071(11 ) .
the nir profiles of these galaxies , however , show features similar to barred galaxies ( `` ib '' ) .
these spirals generally have larger twists compared to the `` ob '' galaxies .
we use shape parameters estimated by the 2mass as the reference .
our analysis should provide estimates similar to the 2mass for contours enclosing larger areas .
since @xmath62 isophote corresponds to a low surface brightness level , i. e. , a region near the edge of a galaxy image , we would expect agreement only in this region . for the purpose of comparison with we use only the relative error between the shape of the contour closest to @xmath3 band @xmath62 isophote and the 2mass estimate .
in most cases we find excellent agreement ( within @xmath94 of the 2mass ) .
however , in a few cases the agreement is poor ( beyond @xmath94 of the 2mass ) and in this section we report only these cases .
we want to stress that the result presented here is obtained after careful comparison of both unsmoothed and smoothed profiles of the parameters . we do not include galaxies whose unsmoothed profiles agree with the 2mass ( i. e. , within @xmath94 of the 2mass ) . for example , we do not include ngc 4394 ( 13 , spiral group 2 ; fig . [ ellps2 ] ) in the list of galaxies given below since its unsmoothed profile shows excellent agreement with the 2mass .
the apparent mismatch between the 2mass and our estimate for smoothed profile is due to the effect of smoothing ( as mentioned in @xmath9 ) .
if the unsmoothed profile of a galaxy shows disagreement ( i. e. , beyond @xmath94 of the 2mass ) in the first place , it is also reflected in its smoothed profile and only then we include the galaxy in the list . for these galaxies , therefore , the estimates of the parameters from both unsmoothed and smoothed profiles disagree with the 2mass .
we note that that an agreement in the @xmath4 profile a galaxy does not necessarily mean that it will also have an agreement in its @xmath5 profile .
note particularly that for several elliptical galaxies we do not have any contour near the @xmath62 level .
we do not attempt to make any comparison for these galaxies .
* ellipticity : * the difference in the estimate of ellipticity is seen only for lenticular and spiral galaxies .
the galaxies showing the difference comprise @xmath95 of the entire sample . here
we provide the list of these galaxies : 11 , 14 in fig .
[ ellpe1 ] ; 5 , 16 in fig .
[ ellpe2 ] ; 1 , 3 , 6 , 8 , 9 , 11 , and 15 in fig .
[ ellps0 ] ; 1 , 6 , 8 14 , and 16 in fig .
[ ellps1 ] ; 1 , 7 , 13 , 14 , and 16 in fig .
[ ellps2 ] ; 3 , 4 , 8 , 11 , 13 , 14 , 15 , and 16 in fig .
[ ellps3 ] ; 2 , 6 , 11 , 13 , 14 , 15 , and 16 in fig .
[ ellps4 ] .
* orientation : * the difference in the estimate of orientation appears in all types of galaxies . for the following galaxies ( @xmath96 of the entire sample ) we have found disagreement : 3 , 8 , 10 , and 13 in fig .
[ ange1 ] ; 6 , 8 , and 9 in fig .
[ angs0 ] ; 2 , 6 and 14 in fig .
[ angs1 ] ; 1 , 7 , 10 , 13 , 14 , and 16 in fig .
[ angs2 ] ; 1 , 8 , 13 , and 15 in fig .
[ angs3 ] ; 1 , 4 , 8 , 9 , 11 , and 15 in fig .
[ angs4 ] . for these galaxies
our estimate of @xmath5 is either lower or higher than the 2mass .
we want to point out that while for some galaxies ( e. g. , 3 and 8 in fig .
[ ellpe1 ] ) the estimates of @xmath4 are in agreement with 2mass , the estimates of @xmath5 for these galaxies ( 3 and 8 in fig . [ ange1 ] ) differ . in some cases our analyses
do not reach the @xmath62 isophote level ( e. g. , 3 and 8 in fig .
[ ellpe1 ] ; 8 and 9 in fig .
[ ellps0 ] ; 6 in fig .
[ ellps1 ] ; 7 in fig .
[ ellps2 ] ) which is a likely reason for disagreement .
we draw reader s attention to galaxies numbered 11 and 14 in fig .
[ ellpe1 ] and listed above .
the @xmath4 and @xmath5 profiles of these galaxies may give an impression to the reader that they do not reach the @xmath62 level ( recall that the figures showing these profiles are for smoothed contours ) .
in fact they do reach the level for the unsmoothed profiles and disagree with the 2mass .
this particular appearance of the profiles arise due to excessive smoothing at the outer part of these galaxies ( see @xmath9 ) .
we provide analyses for smoothed isophotes .
contour smoothing can be thought of as a method to minimize the effect of noise to restore the original shape .
since the 2mass results have not been estimated from smoothed isophotes , we want to emphasize that there is no guarantee that the shape estimates provided by the survey represent the actual shape of galaxy contours . after careful comparison of both unsmoothed and smoothed profiles of galaxies ,
we feel reasonably confident that the disagreement is not entirely due to the methodological difference .
it is likely that our approach is better capable of revealing the actual shapes of galaxy isophotes , especially around the edge .
we have analyzed a sample of galaxies of various hubble types obtained from the 2mass catalogue . the sample contain 112 galaxies imaged in the nir @xmath1 , @xmath2 , and @xmath3 bands .
we have used ellipticity ( @xmath4 ) and orientation angle ( @xmath5 ) , as functions of area within the isophotal contour , as the diagnostic of galaxy shape .
these measures have been constructed from a set of non - parametric shape descriptors known as the minkowski functionals ( mfs ) .
the ellipticity and orientation for each galaxy have been derived at 30 different surface brightness levels in each of these bands .
the ellipticity and position angle ( for @xmath3 band only ) provided by the 2mass are used as the reference in our analysis .
our results show that the elliptical galaxies with @xmath97 appear to be centrally spherical .
these galaxies show smooth variations in ellipticity with the size , quantified by the area within the contour , and are more flattened near the edge than the central region . a variation as strong as @xmath98 , from the center towards the edge , is noticeable in more flattened systems , e. g. , ngc 4125 , ngc 3377 , ngc 4008 , and ngc 5791 ( 9 , 10 , 12 , and 15 in group 2 of elliptical galaxies ; fig .
[ ellpe2 ] ) .
this behavior is similar to previous studies of ellipticals in visual bands ( jedrzejewski 1987 , fasano & bonoli 1989 , franx et al .
the similarity in apparent shapes in optical and nir wavelengths indicates that the low ellipticity noticed at the central regions of these galaxies is intrinsic rather than an artifact contributed by the seeing effect .
additionally the @xmath4 profiles of these galaxies appear very similar in different nir bands with a very low scatter .
this suggests that the morphological differences that are likely to appear in different bands are weak in nir elliptical galaxies . the twist ( @xmath58 ) , characterized by the overall change in the orientation with radius , observed in the orientations of elliptical galaxies decreases with increasing flattening .
a relatively small twist shown by ellipticals with @xmath6 suggests that elongation and isophotal twist are likely to be anti - correlated in the nir wave - bands .
we perform a spearman correlation test between the twist and the deviation in ellipticity ( @xmath57 ) for the entire sample of elliptical galaxies .
the test result ( correlation coefficient -0.35 and probability 0.05 ) indicates a significant anti - correlation between @xmath58 and @xmath57 ( see fig .
[ galletta ] ) .
note that in the test , a small value of the probability with a negative coefficient suggests significant anti - correlation . to check whether this result is due to any methodological artifact or indicates a physical effect
, we rerun the test two more times with different numbers of galaxies in the elliptical sample . for the first run
, we construct the sample after removing only galaxy 10 from group 1 .
this galaxy shows the largest twist ( @xmath99 ) in the entire sample and , therefore , we discard it as an outlier .
we find that the removal of this galaxy slightly weakens the result ( correlation coefficient -0.34 and probability 0.06 ) but the correlation still remains significant . in the next run ,
we remove nine galaxies from the entire sample : the first eight galaxies from group 1 because of their spherical shapes ( @xmath100 ) and galaxy 10 .
this time the test result gives correlation coefficient -0.37 and probability 0.09 .
although a greater probability indicates less confidence for this subsample , the result does not differ substantially from the original sample .
it suggests that the anti - correlation is a physical effect rather than a fluke .
the nir correlation between @xmath58 and @xmath57 for elliptical galaxies is similar to that in the optical wavelengths . for elliptical galaxies in visual bands ,
it has been known for a while that the maximum apparent flattening and the maximum observed twist are inversely related ( galletta 1980 ) .
an interesting aspect of this correlation would be the possible coupling between nir and optical light ( jarrett et al .
2003 ) .
elliptical galaxies with @xmath101 usually show large variation in their orientations .
these galaxies also show considerable differences both in ellipticity and orientation in different bands .
the large twist observed in these galaxies should be taken with caution since spherical contours obtained from nearly spherical galaxies are highly prone to spurious effects such as background noise .
the nir lenticular galaxies have properties similar to those of ellipticals or disk galaxies .
the profiles of a few of these galaxies show trends similar to ellipticals .
a few of these galaxies show characteristic properties resembling to the galaxies with bar like structures .
the lenticular galaxies in our sample , in general , have larger scatter in both @xmath4 and @xmath5 than the ellipticals . in the entire sample of lenticulars ,
at least 2 galaxies have bar like signatures in their profiles .
the observed twists in the lenticulars orientations are comparable to the trend noticed in spherical galaxies ( @xmath101 ) .
the properties of these galaxies stress the fact that morphological classification strongly depends on the wavelength studied
. it may be possible that s0 galaxies do not exist in the long wavelength part of the spectrum .
this is simply our speculation and more elaborate studies are needed to make this a definite conclusion .
the sample of spiral galaxies indicates that a bar like feature is ubiquitous in disk galaxies without any significant dependence on hubble s class .
this sample has 64 galaxies , of which 24 ( @xmath102 ) are optically - classified barred galaxies .
most of these galaxies show features resembling bar(s ) in the infrared .
we label these galaxies as `` optical and infrared bar '' ( `` oib '' ) .
however , we notice something interesting .
all most one - third of these optical barred galaxies lack characteristic features in their nir profiles .
these galaxies include : ngc 4394(12 ) in group 1 ( fig .
[ ellps1 ] ) ; ngc 4375(6 ) and ugc 7073(9 ) in group 3 ( fig .
[ ellps3 ] ) ; ugc 2705(4 ) , ugc 3171(9 ) , ngc 5510(13 ) , and ugc 1478(16 ) in group 4 ( fig .
[ ellps4 ] ) .
the sample of spirals include 40 galaxies that are optically classified non - barred galaxies .
at least 11 of these galaxies show bar like features in their nir profiles , increasing the number of bar like systems to @xmath103 .
several previous studies have also reported a higher frequency of barred systems in disk galaxies ( seiger & james 1998 ; eskridge et al .
2000 ; laurikainen & salo 2002 ) .
our result is in agreement with all of these studies except seiger & james ( 1998 ) , who reported @xmath104 frequency of barred galaxies from a sample of 45 spirals imaged in nir @xmath1 and @xmath105 bands .
our estimate indicates that a significant fraction of disk galaxies are intrinsically non - barred systems .
the absence of a bar like feature in the profiles of 29 spiral galaxies favors this argument .
this is also consistent with eskridge et al .
( 2000 ) . among 19 normal spiral galaxies , 17 galaxies
are sa - type .
the other two , ugc 5739 in group 1 and ic 2947 in group 3 , are irregular / peculiar type galaxies .
in these 17 sa - type spirals , 5 galaxies have distinct bar like features in their shape profiles indicating that @xmath106 of sa - type galaxies have optically hidden bars .
our result is in agreement with eskridge et al .
( 2000 ) who reported @xmath107 galaxies have optically hidden bars from a sample of 186 nir @xmath2-band spirals containing 51 non - barred sa - type galaxies
. note that several disk galaxies show multiple bar features in their profiles .
however , in this study we only report their morphologies and do not attempt to make any definite conclusion regarding their projected structure since identifying galaxies with multiple bars is a complicated issue ( see wozniak et al . 1995 ; erwin & sparke 1999 ) .
it is important to note that our estimate of the frequency of barred galaxies is done by visual inspection . since the sample of galaxies used in this study
is not based on rigorous selection criteria , the results should be taken as an approximate estimate .
the overall isophotal twists observed in the orientations of spiral galaxies range between @xmath108 .
many nir barred spirals appear to have smaller but notable twists in their orientations than the twist at optical wavelengths ( see e. g. , wozniak et al .
two conclusions can be drawn after analyzing the orientation profiles of the nir images .
first , the nir light in @xmath7 and @xmath3 bands is not fully decoupled from population i light .
they are most likely weakly coupled .
second , the different regions ( central bulge , bar , and outer disk ) of disk galaxies are dynamically linked .
it is manifested in the continuous change in the orientations of the majority of the disk galaxies .
exceptions are noticed in few cases , e. g. , ngc 3384 ( 10 , group 1 ; fig .
[ angs1 ] ) , ngc 4152 ( 4 , group 2 ; fig .
[ angs2 ] ) , ugc 2303 ( 1 , group 3 ; fig .
[ angs3 ] ) , where the change in the orientation is very sharp .
but it is important to note that in the region where the sharp changes are observed , the galaxy contours appear to be very spherical ( @xmath109 ) .
a drastic change in the orientations of spherical contours does not indicate reliably the actual nature of the internal structure since these types of contours are prone to spurious effects .
+ + * acknowledgments * we thank the referee for his constructive criticisms and many helpful suggestions .
we are indebted to bruce twarog for thorough and critical reading of the manuscript .
his suggestions improved the paper significantly .
nr thanks tom jarrett , hume feldman , and barbara anthony - twarog for useful discussions .
nr acknowledges the grf support from the university of kansas in 2002 - 2003 .
sfs acknowledges the support from the nonlinear cosmology program , 2003 at observatory de la cote dazur in nice , france during the summer 2003 and the aas travel grant .
this research has made use of the nasa / ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
here we briefly describe the role of contour smoothing in our analyses .
we also discuss the sensitivity of the parameters to any disturbance present on the contour . in order to emphasize the role of contour smoothing we show in fig .
[ ellp_all_e1_unsm ] the ellipticity measured from all three aes , both for unsmoothed and smoothed elliptical galaxies in group 1 . in this figure
the dark , medium , and light colors are used to represent the @xmath7 and @xmath3 band respectively .
the solid , dashed , and dashed - dotted lines show the area , perimeter , and ec ellipses for each band . as we can see from the unsmoothed contours in the left panel , the ellipses differ from one another in each band mostly near the edge .
this deviation is caused by the background noise since the galaxy surface brightness distribution is steeper near the center and shallower outward .
we should also note that in a region far away from the center of a galaxy the absolute value of the distribution is very small . in either case
any kind of background noise will have a strong effect on the outer part of the distribution .
therefore when one constructs contours along the edge of a galaxy , they appear highly deviated no matter what their true shapes are .
if the contours were truly elliptical , they could appear in any arbitrary shape depending on the amount distortion .
since the contours are not elliptic anymore , the aes diverge from one another .
this divergence reflects the apparent shape of the contours .
one can see from the right panel that contour smoothing reduces the role of noise significantly .
it helps restore the original shape of the contours . at the same time
one can also notice that the differences among the ellipses reduce substantially .
to demonstrate further , we collect first four galaxies from each group of spiral galaxies and show the ellipticity for both unsmoothed and smoothed contours in figs .
[ ellp_all_spi_unsm ] and [ ellp_all_spi_sm ] , respectively .
the motive , once again , is to reemphasize the role of contour smoothing . when we compare figs .
[ ellp_all_e1_unsm ] and [ ellp_all_spi_sm ] , it appears that the color difference is , in general , stronger for spiral galaxies and within spirals , it is stronger for certain hubble types .
we see that the galaxies in spiral group 4 , which are mostly late - type , have significant differences in their shapes ( at least what is revealed by the @xmath4 parameter ) when studied in three different nir bands .
we now proceed to check the sensitivity of the minkowski shape measures to detect image contamination by a foreground star .
the @xmath1 band contours of ngc 4278 , ngc 3193 , ngc 3379 ( 1 , 4 , and 7 in elliptical group 1 ) , are shown in fig .
[ star_cont ] .
this figure also includes the contours of ngc 5507 ( 5 in elliptical group 2 ) from the same band . note that the presence of a star is apparent in all three bands .
it is most prominent in the @xmath1 band and this why we show the contour plot in this band .
the left panel of fig .
[ star_measure ] shows the ellipticity profiles of these galaxies and the right panel shows the relative difference in areas enclosed by three tensor aes . in the left panel we use solid , dashed and dashed - dotted lines represent the measures from area , perimeter and ec aes .
the dotted line is the measure from the scalar functional ae that has not been used in this analysis .
it is included only for the purpose of illustration .
note that the construction of the scalar ellipse is different from the tensor ellipses . since the scalar functionals are simply the area ( @xmath26 ) and perimeter ( @xmath27 ) of a given contour
, the scalar ae is an ellipse that has the same area and perimeter as the contour itself . for a perfect elliptical contour ,
the scalar ae will converge with its tensor counterparts .
the scalar functionals do not have enough information to attribute an orientation to its ellipse .
the right panel does not show any measurement from the scalar functional . from figs .
[ star_cont ] and [ star_measure ] , we see that a foreground star embedded in an image causes galaxy isophotes to deviate from their original shapes .
it usually adds small , lobe - like features to the otherwise smoothed and spherical contours .
the functionals easily pick up this type of signal present on the contour and translate it to the shape parameters .
this demonstration highlights the fact that mfs based measures can be used for automatic detection of features attached to the image body .
our analysis is further supported by ngc 5077 ( number 5 in elliptical group 2 ) .
this galaxy is interesting for our purpose since it has been archived in the 2mass catalogue as an image free from foreground star .
the contour plot of the galaxy , however , reveals that it is not the case since the isophotal contours are perturbed by an uncommon feature .
apart from the contour plot , the behavior of the ellipticity profile or the plot showing ratios of the sizes of aes also give strong indication of the presence of something unusual in the image body . from the extent of the feature along the radial direction
one can infer the possible nature of the object that distorts the image contours . a close inspection of the contours and a comparison with the other galaxies mentioned above
lead us to conclude that a foreground star is still left embedded in the galaxy image . with these examples at hand , we feel confident to suggest that both ellipticity and the ratios of the sizes of aes can be used simultaneously as filtering tools in image processing
. these measures may appear useful to reduce the chance of contamination by foreground star while constructing large galaxy catalogues from surveys such as sloan digital sky survey ( sdss ) . | we study a sample of 112 galaxies of various hubble types imaged in the two micron all sky survey ( 2mass ) in the near - infra red ( nir ; 1 - 2 @xmath0 m ) @xmath1 , @xmath2 , and @xmath3 bands .
the sample contains ( optically classified ) 32 elliptical , 16 lenticulars , and 64 spirals acquired from the 2mass extended source catalogue .
we use a set of non - parametric shape measures constructed from the minkowski functionals ( mfs ) for galaxy shape analysis .
we use ellipticity ( @xmath4 ) and orientation angle ( @xmath5 ) as shape diagnostics . with these parameters as functions of area within the isophotal contour , we note that the nir elliptical galaxies with @xmath6 show a trend of being centrally spherical and increasingly flattened towards the edge , a trend similar to images in optical wavelengths .
the highly flattened elliptical galaxies show strong change in ellipticity between the center and the edge .
the lenticular galaxies show morphological properties resembling either ellipticals or disk galaxies .
our analysis shows that almost half of the spiral galaxies appear to have bar like features while the rest are likely to be non - barred .
our results also indicate that almost one - third of spiral galaxies have optically hidden bars .
the isophotal twist noted in the orientations of elliptical galaxies decreases with the flattening of these galaxies indicating that twist and flattening are also anti - correlated in the nir , as found in optical wavelengths .
the orientations of nir lenticular and spiral galaxies show a wide range of twists .
-0.5 in # 1/#1 # 1*#1 * # 1=#1 # 1 # 1 galaxies : morphology - galaxies : structure - galaxies : statistics |
it is generally accepted that quasars are the result of accretion onto massive black holes residing in the nuclei of normal galaxies .
rees ( 1984 ) has argued that a black hole is likely to form at the centre of almost any galaxy , so that the main issue for quasar formation is how the black hole is fuelled .
models for quasar evolution must account for the time dependence of the quasar luminosity function , particularly its peak at @xmath1 and subsequent decline ( boyle , shanks & peterson , 1988 ) , and also for the formation of the massive black holes and their remnants that reside in the nuclei of many nearby galaxies ( ford 1997 ; magorrian 1998 ) . a variety of models has been proposed for the fuelling of quasars , most of which rely on making interstellar gas fall close to the nucleus where it joins an accretion disc ( shlosman , begelman & frank 1990 ) . in this paper
we consider the possibility that the main source of fuel is the hot interstellar medium formed during the collapse of larger galaxies .
provided that the angular momentum of the gas is not too large , a nuclear black hole can grow by bondi accretion and , if the gas temperature is close to the virial temperature , its growth time is controlled largely by the density of the hot gas . in section 3
we show that , soon after the collapse of a protogalaxy , this can be large enough to make the accretion rate of a nuclear black hole comparable to the eddington rate .
it is the protogalaxies with the largest spheroids , large elliptical protogalaxies , that contain most hot gas ( nulsen & fabian 1997 ) , making them the best hosts for quasars in this model .
thus , the observation that quasars reside in elliptical hosts ( mclure 1998 ) is consistent with them being fuelled by hot gas .
based on the presence of companions close to many quasars , it has been argued that gravitational interaction plays a significant role in the quasar phenomenon ( bahcall 1997 ) .
however , we note that this may also be interpreted as indicating recent collapse .
as outlined below , the supply of hot gas is expected to be greatest immediately after a collapse and to decrease with time , so this would also be consistent with fuelling quasars by hot gas .
of course , active galactic nuclei , including quasars , may obtain their fuel from a variety of sources .
we consider that hot gas provides a minimum fuelling rate , which can be supplemented by cold gas , if available .
cooling depletes the hot gas throughout the galaxy , so that the nuclear accretion rate decreases with time after a protogalaxy collapses .
the depletion of the hot gas does not however simply explain the lack of luminous quasars at the current epoch .
the central black holes in nearby elliptical galaxies are still immersed in accretable hot gas , yet generally have low accretion luminosities ( fabian & canizares 1988 ; di matteo & fabian 1997 ) . as an example ,
one of the best candidates for a massive nuclear black hole is m87 ( harms 1994 ; marconi 1997 ) .
a solution to this problem has been argued by di matteo & fabian ( 1997 ) and , specifically for m87 , by reynolds ( 1996 ) in which the accretion flow becomes advection - dominated ( i.e. an adaf , narayan & yi 1995 ) , so having a low accretion efficiency .
we adopt that hypothesis here and assume that when the nuclear accretion rate falls below the threshold for adaf formation most quasars fade rapidly ( fabian & rees 1995 ; yi 1996 ) . in section 4
we describe the incorporation of a simple version of this model for quasar formation into a semi - analytical model for galaxy formation ( nulsen & fabian 1997 ; see also haehnelt & rees 1993 for models relating quasar evolution to galaxy formation ) .
this is used to show that the model can account for the broad features of the history of quasars in section 5 .
section 6 has a brief discussion of feedback on the model , particular that due to compton cooling . in section 7
we discuss the limitations of the semi - analytical model for quasar formation and some predictions of the model .
our conclusions are summarized in section 8 .
shlosman ( 1990 ) discuss the issues of getting a large mass of gas to accrete into the nucleus of a galaxy .
almost all galaxies have appreciable net rotation , so that the main difficulty is the dissipation of angular momentum , essentially all of which must be dissipated in order for gas to accrete into a nuclear black hole .
the total accreted matter should also account for ( most of ) the mass of the nuclear black hole , which exceeds @xmath2 in many cases ( ford 1997 ; magorrian 1998 ) , so that the gas needs to be drained from a large region around the nucleus .
the larger this region , the greater the difficulty of dissipating angular momentum . according to the standard argument , the effective viscosity in an accretion disc can be expressed as ( shakura & sunyaev 1973 ) _
d = s^2 , where @xmath3 is the density and @xmath4 is the isothermal sound speed of gas in the disc , and ( r ) = is the angular frequency of a circular orbit .
the dimensionless parameter @xmath5 can not normally exceed 1 and is generally thought to be @xmath6 0.3 ( e.g. cannizzo 1993 ) .
the same parametrization can be applied to the hot gas , gas at about the virial temperature , but then it is more appropriate to express the viscosity in terms of the scale height , @xmath7 , = s w. the cold gas moves on almost circular orbits , so that the time required for gas from radius @xmath8 to drain to the centre of the disc is roughly ( shlosman 1990 ) = r v_r = r ^2 s^2 , where @xmath9 is the radial speed , @xmath10 is the rotation speed of the gas disc and @xmath11 ( @xmath12 for a keplerian disc ) . since the hot gas is pressure supported , it can drain much faster than a cold disc .
however , angular momentum may still prevent it from accreting directly onto a nuclear black hole .
for gas at about the virial temperature , the scale height is @xmath13 and the speed of sound is close to the kepler speed , @xmath14 , at the same radius . a rough estimate of the time required to dissipate the angular momentum of the hot gas is then .
for @xmath15 of order unity , this is comparable to the dynamical time , while 1 , if the disc gas is cold .
thus , while the drainage of a cold accretion disc is governed by the dissipation of angular momentum , it is only when hot gas flows inward at speeds approaching the free - fall velocity that we need to be concerned with the effects of rotation on the flow .
dissipation of angular momentum is less of a problem for the accretion of hot gas than it is for cold gas . for flow speeds comparable to the speed of sound or faster ,
the angular momentum of the hot gas will be largely conserved , so that its residual angular momentum will cause it to eventually join a disc .
we assume that this occurs at a sufficiently small radius to make the drainage time of the disc from the point where the gas joins it point short .
in the collapse of small protogalaxies , radiative cooling is faster than shock heating , so that gas ends up cold immediately after the collapse ( rees & ostriker 1977 ; white & frenk 1991 ) . in larger systems , which are more tenuous and
have higher virial temperatures , some of the gas can form a hot atmosphere after the collapse .
the condition for the gas at radius @xmath8 to be part of a hot atmosphere is that its radiative cooling time be longer than the free - fall time from @xmath8 to the centre of the protogalaxy .
this cooling time is still significantly smaller than the time at which the system collapses , so that the hot gas will start to cool , forming a cooling flow ( fabian 1994 ) , almost immediately after collapse .
a central black hole can accrete hot gas from the central region of the cooling flow .
based on observations of clusters of galaxies , we expect gas taking part in the cooling flow to be sufficiently inhomogeneous to lead to widespread thermal instability ( nulsen 1986 ; 1988 ) . the general solution for an inhomogeneous cooling flow
is complex , since the flow of each phase must be tracked separately .
however , nulsen ( 1986 ) has argued that gas blobs moving relative to the mean flow tend to be rapidly disrupted until they are small enough to be pinned to the mean flow . as a result ,
the phases tend to flow inward at approximately the same speed , to comove . in the central part of the cooling flow , where conditions change slowly relative to the flow time
, we can also expect the flow to be nearly steady . for the purpose of simulating quasar formation , we take the potentials of galaxies to be exactly isothermal . in that case
, the mean temperature of the inhomogeneous gas mixture in a steady , comoving cooling flow will be close to the constant virial temperature , the exact relationship depending on details of the inhomogeneous density distribution in the gas .
there is a class of self - similar , inhomogeneous cooling flow solutions in which the mean gas temperature is a constant multiple of the virial temperature , the isothermal cooling flows ( nulsen 1998 ) .
the mean gas density and temperature in an isothermal cooling flow are related to the flow time , @xmath16 , by = k kt ( t ) , where @xmath8 is the radius , @xmath17 the flow velocity , @xmath3 the mean gas density , @xmath18 the mean electron number density and @xmath19 the mean hydrogen number density .
@xmath20 is the effective gas temperature ( defined so that the pressure is @xmath21 ) and @xmath22 is the mean mass per gas particle .
the dimensionless constant , @xmath23 , depends on flow details , including the cooling function and the radial dependence of the mass deposition rate . in clusters
the mass flow rate , @xmath24 , is found to be approximately proportional to @xmath8 ( fabian 1994 ; peres et al 1998 ) , which we take to be exact for the purpose of our model . in that case , if the cooling function is approximated by a power law , @xmath25 , then @xmath23 depends weakly on the exponent @xmath26 , ranging from 2.32 to 2.92 for @xmath26 in the range @xmath27 to 0.5 .
we adopt the representative value @xmath28 for our calculations . in an isothermal cooling flow with mass flow rate @xmath29 ,
the flow velocity is constant , regardless of the details of the cooling function , so that the mach number is also constant .
the arguments below rely on the mach number of the cooling flow being close to unity initially , since this maximizes the density of the hot gas in the vicinity of a nuclear black hole .
this is the most critical aspect of the cooling flow model for quasar formation , since it determines the bondi accretion rate of the nuclear black hole . when the mach number of a cooling flow is low , the linear growth of thermal instability is weak ( balbus & soker 1989 ) .
the inhomogeneous , isothermal cooling flow model relies on non - linear thermal instability ( contrary to a common misconception , very small amplitude density fluctuations become non - linear in a cooling flow ; nulsen 1997 ) .
however , for mach numbers of order unity , linear thermal instability is much stronger .
if , contrary to our assumptions , the gas is very nearly homogeneous and the thermal instability is weak ( or if @xmath30 with @xmath31 ) , the mach number of the cooling flow increases inward and , as it approaches unity , thermal instability becomes strong , causing widespread deposition of cold gas .
this tends to make the mach number saturate close to one , so that our assumption that the mach number of the cooling flow in the vicinity of the nucleus is close to one is not sensitive to our assumptions .
a cooling flow with a mach number of order unity is simply a maximal cooling flow .
conveniently , the nuclear accretion rate can be determined exactly for the isothermal cooling flow model ( see below ) , but our quasar formation model is not critically dependent on the assumption that the hot gas forms an isothermal cooling flow . also note that the gas density in an isothermal cooling flow with mach number close to one is similar to that obtained by other arguments for the maximum gas density in a protogalaxy ( fall & rees 1985 ) .
we assume that a nuclear black hole accretes any hot gas coming within its influence by bondi accretion . at very small
@xmath8 the residual angular momentum causes the accreting gas to pass through a shock ( assumed to be radiative ) and join an accretion disc .
thus , the final stages of the accretion are still assumed to be through a disc , but we take the nuclear accretion rate to equal the bondi rate .
thus , the nuclear accretion rate is determined by the density and temperature of the hot gas at the point that the influence of the black hole becomes dominant .
there are a number of reasons why this assumption may not be valid , but we have adopted it as the simplest possibility .
the result ( [ igc ] ) shows that the steady cooling flow is governed by the requirement that the cooling time equals the flow time , to within factors of order unity .
as cooling gas comes under the influence of the black hole , its flow velocity will increase , reducing @xmath16 to the point that cooling is no longer effective .
thus , the transition between the cooling flow and the bondi solution occurs at about the radius where the initial mach numbers of the two flows are equal .
the bondi accretion rate for a monatomic gas ( with @xmath32 ) is ( shu 1991 ) = , where @xmath33 is the gas density and @xmath34 the adiabatic sound speed at large @xmath8 , and @xmath35 is the mass of the black hole .
well outside the accretion radius , the density in the bondi solution is almost constant so that the velocity , @xmath36 , can be determined from the accretion rate , @xmath37 .
the gas temperature is also nearly constant , so that mach number is given approximately by = ( g2 ^ 2 r)^2 . equating this to the mach number , @xmath38 , of the cooling flow
, we find that the bondi solution takes over at about the radius = g2 ^2 ^1/2 . using ( [ igc ] ) for the gas density and ( [ rx ] ) to replace @xmath8 in the result , we can evaluate the accretion rate ( [ bondi ] ) as = 2 k ^3/2 kg ( ) ^2 , where @xmath39 is the gas temperature at large @xmath8 ( note that the last factor is a constant ) .
this result is shown to be exact in the appendix , where @xmath39 is now to be interpreted as the `` mean '' gas temperature of a steady isothermal cooling flow and @xmath38 as its mach number .
the argument in the appendix also shows that , for the conditions of our model , gas conditions that would give an isothermal cooling flow with @xmath40 , the nuclear accretion rate is independent of the gravitational potential in between the edge of the steady cooling flow and the nucleus .
thus , the accretion rate is largely unaffected by changes to the potential of the galaxy due to deposition of cooled gas .
this feature of the model reflects an accidental balance between the competing effects of deepening the potential : a temperature rise , reducing the bondi accretion rate , and a density increase , tending to increase it .
the result ( [ mdot ] ) shows that a black hole at the centre of a protogalaxy grows exponentially by bondi accretion .
the timescale for growth is = = 12 k ^2 ( ) k g ^3/2 , which depends only on the gas temperature and the mach number in the outer parts of the of the steady cooling flow .
numerically , 5.110 ^ 8 _ -23 t_6 ^ -1 ^-3/2 y , where @xmath41 k and @xmath42 . this can be compared directly with the growth timescale of a black hole accreting at the eddington rate , @xmath43 , as 11 _
-23 t_6 ^ -1 ^-3/2 t_edd .
thus , for the relevant temperatures , the accretion rate from a maximal cooling flow is about one tenth or more of the eddington rate .
the bondi radius for a stellar mass object is very small , = g^2 610 ^ 11 t_6 ^ -1 cm , for @xmath35 in solar masses , so we should be wary of applying our result to accretion onto very small black holes .
however , we assume that more massive seed black holes are formed in the cores of all galaxies , as described by rees ( 1984 ) , and that these objects then grow to quasars by the accretion of hot gas .
in order to test the outcome of this quasar formation model , we have incorporated it into a semi - analytical model for galaxy formation , the details of which are are described in nulsen & fabian ( 1997 ) and nulsen , barcons & fabian ( 1998 ) . in this section
we outline modifications we have made to that model in order to track the growth and accretion luminosities of nuclear black holes .
as outlined in the previous section , the major factors that determine the growth rate of a nuclear black hole are the gas temperature and the mach number of the cooling flow .
expressed in terms of the growth time ( [ growth ] ) , the time dependence of the mass of a nuclear black hole is given by = _
t_1^t_2 dt .
the growth time depends on the temperature of the gas , @xmath39 , which is determined at the time of collapse of a protogalaxy , and the mach number , @xmath38 , of the isothermal cooling flow .
the latter factor is the only time dependent part of @xmath44 . in our galaxy formation model ,
each collapse produces a dark halo which is taken to be a perfect isothermal sphere ( density @xmath45 ) that is truncated at @xmath46 , the radius within which the mean density is 200 times the background density of an einstein - de sitter universe at the time of collapse .
the gas temperature produced in the collapse is expressed as = ^2 k , where @xmath47 is the line - of - sight velocity dispersion of the halo and @xmath48 is a dimensionless parameter , generally lying in the range 0.5 to 1 , that allows for excess energy in the gas ( mostly excess binding energy resulting from supernova driven ejection ; @xmath49 corresponds to zero excess energy ; @xmath48 here is determined for each collapse as in the existing semi - analytical model ) .
the outcome of a collapse is determined by considering a notional , non - radiative collapse . in this collapse ,
the gas would form a hydrostatic atmosphere , with density proportional to @xmath50 ( also truncated at @xmath46 ) .
the ratio , @xmath51 , of the cooling time to the free - fall time in the notional atmosphere is used to separate the gas into two parts , one with @xmath52 that is cold ( due to efficient radiative cooling ) immediately after the collapse and one with @xmath53 that forms a hot atmosphere after the collapse .
the model parameter @xmath54 is of order unity and determines the radius , @xmath55 , in the notional atmosphere that separates these two regions .
radiative cooling eventually causes hot gas produced in the collapse to cool to low temperatures . as in clusters ,
the hot gas is little affected by cooling until the age of the system is comparable to its cooling time , at which stage it joins a steady cooling flow before being deposited as cold gas .
thus the hot atmosphere consists of an outer region that is largely unperturbed since the collapse , a transition region , comparable in size to the steady cooling flow , and a central steady cooling flow .
the total rate of mass deposition and the extent of the steady cooling flow are determined by the initial state of the hot atmosphere ( fabian & nulsen 1979 ) .
in particular , this means that the mach number of the steady cooling flow is determined by the initial structure of the hot gas .
formerly , the hot gas was assumed to cool to low temperature at a time @xmath56 , where @xmath57 is time of the collapse and ( r ) = 3(r ) k 2 ( r)(r ) ( ) r^-2 is the cooling time of the hot gas in the notional collapse .
this made deposition of the cooled gas start discontinuously a short time after collapse . in reality
, the onset of the cooling flow will be continuous , commencing immediately after the collapse . to model
this , here we assume that the gas cools when t = + ( r ) - ( ) , which gives the radius ( in the notional atmosphere ) of gas that is cooling at time @xmath58 as = ^1/2 .
since the cooling time of the first hot gas to cool is comparable to the free - fall time ( @xmath54 about 1 ) , we should expect the initial mach number of the cooling flow , @xmath59 , to be about 1 .
we treat this as a parameter of the model .
the time dependence of the mach number is then determined from the expectation that the flow velocity scales with time as @xmath60 , giving = ^(1 - 2)/2 .
using this and ( [ growth ] ) in ( [ mass ] ) enables us to determine the factor by which the mass of a nuclear black hole grows as a function of the time .
differentiating the result with respect to the time gives the accretion rate and hence the luminosity of the black hole .
of course , the black hole stops growing when the hot gas is exhausted ( @xmath61 ) .
the low level of emission from nuclear black holes in nearby galaxies suggests that the radiative efficiency of accretion is much lower now than it was in quasars ( fabian & rees 1995 ; di matteo 1999 ) .
several models , including advection - dominated accretion flows ( narayan & yi 1995 ) , advection - dominated inflow - outflow solutions ( blandford & begelman 1999 ) and other gas processes ( stone , pringle & begelman 1999 ) , suggest that this is due to a reduced gas supply . despite indications that advection - dominated accretion discs do not account for the behaviour of some nearby massive black holes ( di matteo et al 1999 ) , for the sake of definiteness
, we base our model on them .
thus the radiative efficiency is assumed to be high as long as the accretion rate exceeds about @xmath62 ( esin , mcclintock & narayan 1997 ) .
when the accretion rate falls below this value , the radiative efficiency of the nuclear accretion disc is assumed to plummet , in effect turning off emission from the active nucleus .
this gives a critical growth time , = 1.3 ^2 3.5 10 ^ 7 ^-2 y , and when the growth time ( [ growth ] ) exceeds @xmath63 , a quasar turns off , although the nuclear black hole can continue to grow .
@xmath5 is treated as a parameter of the model .
our galaxy formation model uses the block model ( cole & kaiser 1988 ) to simulate merger trees .
the smallest blocks have a mass of @xmath64 . in order to simulate the presence of seed black holes , a black hole of ( arbitrary ) unit mass
is associated with each of the smallest blocks .
when a block collapses , the black holes associated with all merging sub - blocks are assumed to merge into a single black hole .
this makes the mass of a seed black hole proportional to the mass of its halo up to the stage that it starts to grow by bondi accretion .
a nuclear black hole can only grow by bondi accretion when it forms in a collapse that produces some hot gas .
such systems are identified with normal galaxies in our model .
the model does not allow mergers between normal galaxies , so that normal galaxies only grow in collapses where they accrete dwarf galaxies and gas .
any collapse involving more than one normal galaxy is taken to form a group or cluster of galaxies . since
nuclear black holes are associated with galaxies rather than a group or cluster , we do not track the growth of black holes for galaxies in these systems .
in short , black holes can only grow by bondi accretion to form quasars in the `` normal '' galaxies of our simulation .
mcclure et al ( 1998 ) find from optical imaging that most quasars do occur in elliptical , or spheroidal galaxies , so our model should apply well to such objects .
it may not be directly relevant to present - day low luminosity seyfert galaxies , which tend to be in spiral galaxies , but can have undergone a hot - phase era at an earlier stage .
note that , since the black hole masses are in arbitrary units , the quasar luminosities are too . adopting a luminosity scale and radiative efficiency for the quasars will fix the scale of the black hole masses . on the bolometric magnitude scale used in the plots in this paper
, @xmath65 corresponds to an accretion rate of @xmath66 black hole mass units per year .
if the radiative efficiency is 0.1 and the mass unit is taken as @xmath67 , this would correspond to a bolometric luminosity of about @xmath68 .
for the purpose of the simulations we have taken an open cdm cosmology , with @xmath69 , density parameter @xmath70 , baryon density parameter @xmath71 and @xmath72 . because nuclear black holes grow exponentially in our model , the results are quite sensitive to the key parameters , @xmath54 , @xmath59 and @xmath5 . using parameter values that favour high growth produces such massive nuclear black holes by the present day that the most recent quasars are inevitably the most luminous . at the other extreme
, parameter values can easily be found that result in essentially no growth of the seed black holes .
the range of parameters giving substantial , but not excessive , black hole growth is relatively narrow ( although it covers a substantial part of the physically reasonably parameter range due to the correlated effects of the parameters ) .
models presented here are chosen to lie in that range .
1 shows distributions of total accretion luminosity for several redshifts for the case @xmath73 , @xmath74 and @xmath75 , while fig .
2 shows the same thing for @xmath76 .
since @xmath5 only affects the the critical growth time ( equation [ critgrow ] ) , black hole masses and accretion rates are identical in the two models .
differences between them are entirely due to the earlier onset of the adaf phase for the model of fig .
2 . this effect is greatest at low redshifts , since a greater proportion of the active nuclei is then old and cooling flows in the older collapsed systems have lower mach numbers ( equation [ macht ] ) .
a black hole accreting from a cooling flow with a low mach number has a longer growth time ( equation [ growth ] ) and so is more likely to be an adaf when @xmath5 is increased .
this accounts for the substantial reduction in the numbers of luminous active nuclei at low redshifts between the models of fig .
1 and fig . 2 .
the model in fig .
3 is the same as that of fig . 1
, except that the initial mach number of the cooling flow is @xmath77 .
increasing the mach numbers of the cooling flows increases the nuclear accretion rate , resulting in greater black hole growth and higher nuclear luminosities .
this effect can be seen in fig . 3 , where , for the same seed mass , the most luminous quasars are more numerous at all redshifts .
finally , fig .
4 shows the bolometric luminosity distributions for a model with @xmath78 , @xmath77 and @xmath79 .
this is most readily compared to the model of fig .
the effect of changing @xmath54 is more complicated than that of the other two parameters , but its main influence on black hole growth is through the cooling time at the inner edge of the notional hot atmosphere . @xmath54 sets the ratio of the cooling time to the free - fall time at @xmath55 , the inner edge of the notional hot atmosphere , so that reducing it reduces the cooling time there , @xmath80 .
this cooling time sets the timescale for the evolution of the mach number of the cooling flow ( equation [ macht ] ) .
reducing @xmath80 causes the mach number of the cooling flow to decrease more quickly , reducing the overall growth of the nuclear black holes and hence their luminosities .
the bolometric luminosity is , essentially , just the total accretion rate of the black holes and not likely to be a good measure of the visible luminosity of the disc . despite its shortcomings , for the sake of definiteness
, we use the thermal disc model ( shakura & sunyaev 1973 ) to estimate the visible luminosity .
this gives the emitted spectrum p_^1/3 ^2/3 ^2/3 , for frequency @xmath81 .
the resulting `` visible '' luminosity functions are plotted in fig . 5 , for the quasar formation model of fig . 1 ( @xmath73 , @xmath74 , @xmath79 ) . to compare these to the observed quasar luminosity functions ( boyle 1988 ) , we need to convert our arbitrary magnitude scale to absolute blue magnitude .
based on the intensity spectrum of the x - ray background , fabian & iwasawa ( 1999 ) argue that 85 percent of the accretion power of quasars is absorbed ; only ten per cent is seen without some obscuration at 1 kev
. if so , then the number densities in fig . 5 ( and preceding figs ) should be reduced by about a factor 10 . in that case , adding 2 3 to our arbitrary magnitudes to convert to @xmath82 gives rough agreement between our number densities and those of boyle ( 1988 ) . with this conversion ,
our luminosity functions are too flat below about @xmath83 and probably too steep above .
the sharp cut offs in the model luminosity functions are largely due to our assumption that the mass of a seed black hole is proportional to the mass of the dark halo in which it forms . a more realistic model for the seed black holes would give them a distribution of masses . in effect ,
this distribution would be convolved with the luminosity functions , making them fit the observed luminosity functions better .
the redshift dependence of our model is also only in rough agreement with the observed quasar luminosity functions .
the greatest discrepancy is the absence in our model of luminous active nuclei at @xmath84 and earlier .
while this is affected to some extent by the collapse model ( cosmology ) , in large part it is due to the time required to grow massive black holes . from equation ( [ growth ] ) , the growth rate is maximized by minimizing @xmath85 , which generally means in the hottest collapses .
the virial temperature of a halo of mass @xmath86 collapsing at time @xmath57 scales as @xmath87 , so that , for a given mass , the earliest collapses give the most growth .
however , few massive galaxies collapse early , and , since it requires several growth times ( equation [ ngrow ] ) to produce a massive black hole , very few of these form early in our model .
6 is a contour diagram of the distribution of the masses of the nuclear black holes vs spheroid mass at @xmath88 in the model of fig . 1 . magorrian ( 1998 ) and richstone ( 1998 ) find that black hole mass is proportional to blue luminosity of the spheroid ( bulge ) .
however , there is a substantial spread in this relationship and the data would be consistent with a substantially steeper relationship for the more massive bulges .
this is roughly consistent with the ridge and the lower spur at high spheroid mass of the distribution shown here .
however , there is no evidence in the data for the extension to high black hole mass for spheroids of about @xmath89 in the model . taking @xmath90 black hole mass units to correspond to a spheroid mass of @xmath91 and using @xmath92 ( richstone 1998 ) gives the conversion factor of @xmath93 per black hole mass unit . in that case
, a bolometric magnitude of @xmath65 in the figs would correspond to a bolometric luminosity of about @xmath94 .
we can also compare the results of fig . 1 to the quasar x - ray luminosity function ( miyaji , hasinger & schmidt 1998 ) . using the black hole mass calibration from above and a fixed bolometric correction of about 50 for the 0.52 kev x - ray luminosity
( elvis 1994 ; fabian & iwasawa 1999 ) means that a magnitude of @xmath65 in fig . 1 corresponds to a 0.52 kev x - ray luminosity of about @xmath95 .
assuming that absorption reduces the x - ray luminosity function by a factor of about 10 , as above , there is rough agreement between the results in fig . 1 and the observed x - ray luminosity function of miyaji ( 1998 ) .
however , the fit suffers from essentially the same problems that we found for the visible luminosity function . as in that case
, the most serious problem is the lack of high redshift quasars .
ciotti & ostriker ( 1997 ) argue that feedback from a quasar will stifle a cooling flow by heating the cooling gas . by raising the gas temperature and reducing its density , this could also dramatically reduce the bondi accretion rate .
however , using ferland s ( 1996 ) cloudy and the quasar spectra of laor et al ( 1997 ) , we find that the compton temperature of radio - loud and radio - quiet quasars is approximately one and two million k , respectively .
this is cooler than the virial temperature of most of the systems that form quasars .
furthermore , the gas temperature rises inward in the bondi accretion flow , so that compton feedback from the quasar is likely , at best , to cool the accreting gas .
if compton cooling is significant , then it will almost certainly increase the nuclear accretion rate , possibly causing it to approach the eddington limit ( fabian & crawford 1990 ) .
if compton cooling reduces the temperature of gas near to the nucleus , the reduction in pressure will result in inflow at speeds comparable to the sound speed in the uncooled gas .
since the flow within the bondi radius is already roughly sonic , the accretion rate will not be altered dramatically unless compton cooling is effective beyond the bondi radius .
taking the compton cooling time as @xmath96 , where @xmath97 is the electron mass , @xmath98 is the thomson cross section and @xmath99 is the nuclear luminosity , and taking the flow time from the bondi radius , @xmath100 , as @xmath101 , we have = 3c^2 gl_h = 3 c l_edd 4 l_h 0.8 t_6 ^ -1/2 ( l_h l_edd)^-1 , where the nuclear luminosity has been put in terms of the eddington luminosity , @xmath102 .
this shows that compton cooling will be significant outside the bondi radius for most eddington - limited active nuclei . in terms of our model , for the accretion rate ( [ mdot ] ) , if the radiative efficiency of the nuclear accretion disc is @xmath103 ( and @xmath104 ) , then 9 _
-1 ^ -1 ^-3/2 t_6 ^ -3/2 _ -23 .
we find that the ionization parameter of the gas at the accretion radius , @xmath105 under these circumstances , nuclear radiation keeps the gas highly photoionized , so that the effective cooling function is close to pure bremsstrahlung .
the compton cooling time is then less than the infall time for gas temperatures exceeding about @xmath106 k when the mach number @xmath107 .
this means that the accretion rate may well exceed the bondi rate in the cases when it would be highest .
no allowance has been made for this in our model . a further effect which may influence some objects is feedback due to radio jets .
if the central engine produces jets or outflows which deposit significant energy near the bondi radius , then the accretion rate can be much reduced .
it is not clear how such an effect should be included in our model at this stage . if , as suggested by mclure ( 1998 ) , the radio loud quasars are those with the largest black holes , then feedback from radio jets might be responsible for limiting the growth of the black holes in these systems . finally , it has also been suggested that a wind might expel the surrounding gas when quasars becomes sufficiently luminous ( silk & rees 1998 ; fabian 1999 ) .
this would lead to a much closer correlation between bulge and remnant mass .
the implementation of the quasar formation model used here has a number of shortcomings .
first , we only follow the growth of black holes in isolated galaxies .
this discounts growth in groups and clusters .
because of the high gas temperature , a central galaxy in a group could potentially accrete very rapidly .
however , the block model gives no information about the spatial arrangement of collapsing objects , so we are unable to identify central galaxies in groups and clusters .
we may therefore be ignoring the most luminous quasars and the most massive black holes .
the truncated isothermal potentials used in the model lead to gas density distributions that are more peaked than in more realistic collapse models ( navarro , frenk & white 1997 ) .
this affects the time development of the cooling flows , changing the evolution of the mach number , and so would affect the time dependence of the nuclear accretion rate ( equation [ mdot ] ) .
however , the cooling time of the hot gas is comparable to the free - fall time in normal galaxy collapses , so we should still expect immediate onset of a cooling flow with initial mach number close to 1 in most cases .
the initial cooling time controls the rate of change of the mach number while it is close to 1 ( when the growth rate is largest ) and this is comparable to the collapse time .
thus , we do not expect such a change to have a dramatic effect on the results of the simulation . beyond this ,
is not clear how a more realistic collapse model would alter our results . in our simple cooling flow model ,
the nuclear accretion rate is insensitive to details of the galactic potential .
however , this may change in a more realistic model , such as one in which a central star cluster is formed . in that case ,
matter deposited by the cooling flow beyond the bondi radius could significantly alter the central potential and so affect the nuclear accretion rate .
the handling of abundances in our galaxy formation model is very crude , only allowing for the effects of type ii supernovae and treating the gas as homogeneous .
increasing the abundance increases the cooling function , hence the growth time ( equation [ growth ] ) , and so would reduce black hole masses and quasar luminosities .
this may be significant , since the abundances in some quasars appear to be very high ( hamann & ferland 1993 ; ferland 1996 ) .
on the other hand , the cooling function is considerably less sensitive to abundance for temperatures exceeding about @xmath106 k and the gas temperature exceeds this value in most of the systems that would be quasars , so we should not expect this to have a major effect on the outcome of the model . as discussed in section [ results ] , our assumption that the mass of a seed black hole is proportional to the mass of halo in which it resides is too simplistic . given that seyfert nuclei can occur in disc galaxies , it seems likely that some active nuclei are not fuelled by hot gas ( or , at least , not by gas from a hot halo resulting from the collapse of the protogalaxy ) .
a wide variety of other mechanisms for fuelling active nuclei have been proposed , including starburst activity , interactions between galaxies and the effects of a bar .
there is also some cold gas within the region that is effectively drained through a cold accretion disc .
some or all of these gas sources may fuel seed black holes , in which case , they could have a wide range of masses , depending on details of the history of each galaxy .
such effects would be compounded with those due to the processes described in rees ( 1984 ) , that are also likely to lead to a range of seed masses .
if a large proportion of active nuclei are heavily absorbed , then mergers between galaxies may affect their luminosity by disturbing the absorbing material , which could alter the luminosity in either direction .
in other words , it is possible that much of the evolution of observed optical quasars is due to changes in the obscuring material rather than changes in accretion rate .
the worst shortcoming of our model is its failure to produce quasars at high redshifts . for the cosmological parameters used
, a @xmath108 halo would have collapsed from a @xmath109 peak at about @xmath110 y ( @xmath111 ) .
the gas temperature in a halo of mass @xmath112 , collapsing at @xmath113 y , is t 3.310 ^ 6 m_12 ^ 2/3 t_9 ^
-2/3 ^-1 k , for @xmath48 as defined in equation ( [ beta ] ) , so that the growth time for a nuclear black hole in such a system would have been 1.510 ^ 8 ^-3/2 m_12 ^ -2/3 t_9 ^ 2/3 _
-23 y. the cooling function depends on abundances and ionization ( bhringer & hensler 1989 ) , but @xmath114 lies in the the range 0.5 2 for the relevant temperatures .
thus , an early collapsing protogalaxy would have had roughly 10 @xmath115-folding times to form a massive nuclear black hole before @xmath116 .
this shows that the lack of high redshift quasars is not a fundamental shortcoming of our model , although details of the model would clearly need to be modified in order to account for them .
the failure of the model to account for the observed relationship between black hole and spheroid mass ( richstone 1998 ) is due primarily to the exponential black hole growth , which tends to make the larger black holes grow very large .
as discussed in the previous section , different feedback mechanisms may enhance or limit the growth rate .
if feedback from a radio source limits the growth of the most massive black holes , then this , rather than the fuel supply , might be the cause of the relationship between spheroid and black hole mass .
note that we have ignored many effects in the nuclear accretion disc that could break the simple connection we have assumed between the bondi accretion rate and the nuclear accretion rate .
finally , we have attempted to account for the masses of the remnant nuclear black holes and the evolution of the quasars with a single mechanism .
if the processes that form the seed black holes account for a substantial part of their mass , or , if other gas sources also play a significant role in the fuelling of active nuclei , then this will not be possible .
in that case , accretion of hot gas may simply be one of several fuel sources for quasars .
nevertheless , the results in section [ feed ] show that hot gas is potentially a significant fuel source for agns .
conditions are most favourable for quasar formation in our model when the hot gas supply is greatest , soon after the collapse of a large protogalaxy . in hierarchical collapse models ,
a collapse will generally include the infall of gas and other galaxies , so that the presence of a close companion of comparable luminosity may be interpreted as an indication of recent collapse .
thus , the results that have been interpreted as showing the gravitational interactions can drive quasars ( bahcall 1997 ) may also be interpreted as indicating that quasars form in systems that have collapsed recently , as expected in the present model . for this purpose ,
the main difference between the predictions of the models is that no close companion is required in the case that quasars are fuelled by hot gas .
if quasars are fuelled by hot gas , then there should be substantial halos of hot gas around them .
since the gas temperature typically exceeds @xmath106 k , these should be detectable by their x - ray emission .
the limited extent of the hot gas ( comparable to the size of the dark halo ) makes it hard to separate from the powerful nuclear x - ray emission of a quasar .
nevertheless , the angular resolution of chandra should be sufficient to detect diffuse emission around some quasars out to redshifts of about 1 .
bondi accretion of the hot gas produced in the collapse of protogalaxies onto a seed population of nuclear black holes is sufficient to form and fuel quasars .
a simple simulation shows that this model can account for the optical and x - ray luminosity functions of quasars for @xmath117 , provided that about 90 percent of quasars are obscured .
the simulation produces insufficient quasars at high redshifts and predicts a wider range of black hole masses in massive spheroids than has been found .
hot gas formed in the collapse of large protogalaxies is likely to be the minimum source available for fuelling quasars and so can form the baseline above which other sources contribute .
our model directly confronts and includes problems related to the current fuel supply of massive black holes in elliptical galaxies .
the details of the amplitude and evolution of hot gas as a fuel supply is sensitive to the presence of plausible feedback mechanisms , such as heating due to radio jets .
the bondi accretion rate from the hot gas formed in the collapse of a protogalaxy can exceed the eddington accretion rate in systems with virial temperatures exceeding about @xmath106 k , and can be enhanced by feedback due to compton cooling in such systems .
thus , hot gas is an excellent fuel source for quasars .
pejn gratefully acknowledges the hospitality of the institute of astronomy , cambridge , during part of this work .
acf thanks the royal society for support .
balbus s. a. , soker n. , 1989 , apj , 341 , 611 bahcall j. n. , kirhakos s. , saxe d. h. , schneider d. p. , 1997 ,
apj , 479 , 642 blandford r. d. , begelman m. c. , 1999 , mnras , 303 , l1 bhringer h. , hensler g. , 1989 , aa , 215 , 147 boyle b. j. , shanks t. , peterson b. a. , 1988 , mnras , 235 , 935 cannizzo j.k .
, 1993 , apj , 419 , 318 ciotti l. , ostriker j. p. , 1997
, apj , 487 , l105 cole s. , kaiser n. , 1988 , mnras , 233 , 637 di matteo t. , fabian a. c. , 1997 , mnras , 286 , l50 di matteo t. , fabian a. c. , rees m.j . , carilli c.l . ,
ivison r.j . , 1999 ,
mnras , 305 , 492 elvis m. , wilkes b. j. , mcdowell j. c. , green r. f. , bechtold j. , willner s. p. , oey m. s. , polomski e. , cutri r. o. c. , 1994 , apjs , 95 , 1 esin a. a. , mcclintock j. e. , narayan r. , 1997 , apj , 489 , 865 fabian a. c. , 1994 , araa , 32 , 277 fabian a. c. , 1999 , mnras , submitted fabian a. c. , canizares c. r. , 1988 , nat , 333 , 829 fabian a. c. , crawford c. s. , 1990 , mnras , 247 , 439 fabian a. c. , iwasawa k. , 1999 , mnras , 303 , l34 fabian a. c. , nulsen p. e. j. , 1979 , mnras , 186 , 783 fabian a. c. , rees m. j. , 1995 , mnras , 277 , l55 fall s. m. , rees m. j. , 1985 , apj , 298 , 18 ferland g. j. , 1996 , hazy , a brief introduction to cloudy , univ kentucky dept physics & astronomy internal report ferland g. j. , baldwin j. a. , korista k. t. , hamann f. , carswell r. f. , phillips m. , wilkes b. , williams r. e. , 1996 , apj , 461 , 683 ford h. c. , tsvetanov z. i. , ferrarese l. , jaffe w. , 1997 , iau symposium 184 kyoto 1997 , the central region of the galaxy and galaxies , astro - ph 9711299 haehnelt m. g. , rees m. j. , 1993 , mnras , 263 , 168 hamann f. , ferland g. , 1993 , apj , 418 , 11 harms r. , ford h. , tsvetanov z. , hartig g. , dressel l. , kriss g. , bohlin r. , davidsen a. , margon b. , kochhar a. , 1994 , apj , 435 , l35 laor a. , fiore f. , elvis m. , wilkes b. j. , mcdowell j. c. , 1997 , apj , 477 , 93 magorrian j. , , 1998 , aj , 115 , 2285 marconi a. , axon d. j. , macchetto f. d. , capetti a. , sparks w. b. , crane p. , 1997 ,
mnras , 289 , l21 mclure r. j. , dunlop j. s. , kukula m. j. , baum s. a. , odea c. p. , hughes d. h. , 1998 , apj , submitted , astro - ph 9809030 miyaji t. , hasinger g. , schmidt m. , 1998 , proceedings for highlights in x - ray astronomy in honour of joachim trmper s 65th birthday , astro - ph 9809398 narayan r. , yi i. , 1995 , apj , 452 , 710 navarro j. f. , frenk c. s. , white s. d. m. , 1995 , mnras , 275 , 56 nulsen p. e. j. , 1986 , mnras , 221 , 377 nulsen p. e. j. , 1988 , in a. c. fabian ed , cooling flows in clusters and galaxies , kluwer , dordrecht , 175 nulsen p. e. j. , 1997 , in n. soker ed , galactic and cluster cooling flows , astronomical society of the pacific , san francisco , 135 nulsen p. e. j. , 1998 , mnras , 297 , 1109 nulsen p. e. j. , fabian a. c. , 1997 , mnras , 291 , 425 nulsen p. e. j. , barcons x. , fabian a. c. , 1998 , mnras , 301 , 168 peres c.l . , fabian a.c . , edge a.c . ,
allen s.w . ,
johnstone r.m .
, white d.a .
, 1998 , mnras , 298 , 416 rees m. j. , 1984 , araa , 22 , 471 rees m. j. , ostriker j. p. , 1977 ,
mnras , 179 , 541 reynolds c. s. , di matteo t. , fabian a. c. , hwang u. , canizares c. r. , 1996 , mnras , 283 , l111
richstone d. , , 1998 , nat , 395 , a14 shakura n. i. , sunyaev r. a. , 1973 , aa , 24 , 337 shlosman i. , begelman m. c. , frank j. , 1990 , nat , 345 , 679 shu f. h. , 1991 , the physics of astrophysics , volume ii , university science books , mill valley silk j. , rees m. j. , 1998 , a&a , 331 , l1 stone j. d. , pringle j. e. , begelman m. c. , 1999 , mnras , in press white s. d. m. , frenk c. s. , 1991 , apj , 379 , 52 yi i. , 1996 , apj , 473 , 645
we consider bondi accretion by a massive black hole at the centre of a steady inhomogeneous cooling flow .
we will show that , for the cooling flow model used here , the accretion rate onto the black hole depends on conditions at the edge of the steady cooling flow , but is insensitive to details of the gravitational potential between there and the central black hole .
in particular , when the ratio of specific heats , @xmath118 , the accretion rate is completely independent of the intervening potential .
the cooling time of the gas is = p ( -1 ) ( t ) , where @xmath119 is the pressure , @xmath18 the electron density , @xmath19 the hydrogen ( proton ) number density , @xmath20 the temperature and @xmath120 the cooling function .
the flow time of the gas is = r v , where @xmath8 is the radius and @xmath17 the flow velocity ( positive inward ) . for the bondi solution ( shu 1991 ) at small @xmath8 , @xmath121 ,
the density varies as @xmath122 and the temperature as @xmath123 , so that the ratio of the cooling time to the flow time varies as ~t(t ) , since the temperature increases as @xmath8 decreases and , for the relevant temperatures and abundances , @xmath124 is almost always an increasing function of @xmath20 , this ratio increases with decreasing @xmath8 . in a steady cooling flow @xmath125 ( fabian 1994 ) , but as the flow comes under the influence of a central black hole @xmath126 will increase , eventually making cooling negligible .
thus , at sufficiently small @xmath8 cooling can be ignored and the flow asymptotes to the bondi solution .
we begin by outlining the bondi solution . for a steady , spherical flow ,
the mass flow rate ( inward ) is m = 4v r^2 = constant , where @xmath3 is the gas density .
the flow is adiabatic , so that = t_0 _ 0 ^ -1 , where @xmath127 and @xmath128 are evaluated a long way from the black hole . using these
, we can express the flow speed as v = m 4_0 r^2 ( t_0t)^1/(-1 ) .
bernoulli s theorem gives h + v^2 - gmr = h_0 , where @xmath8 is the radius , @xmath129 $ ] is the specific enthalpy of the gas , @xmath130 is the specific enthalpy at temperature @xmath127 and @xmath86 is the mass of the black hole . differentiating ( [ benerg ] ) with respect to @xmath8 and using ( [ vbondi ] ) to find @xmath131 gives ( h - v^2 - 1 ) 1tdt dr = 2v^2 r - gmr^2 , so that at the sonic point , @xmath132 , where @xmath133 , we must have v_s^2 = gm2 r_s . using these results in ( [ benerg ] )
gives = 4 ( -1 ) 5 - 3 h_0 , enabling us to evaluate all quantities at the sonic point in terms of @xmath127 , @xmath128 and @xmath86 .
evaluating the mass flow rate at the sonic point then gives the bondi accretion rate as = _ 0 ( gm)^2 s_0 ^ 3 ( 5 - 32)^3 - 5 2(-1 ) , where @xmath134 is the speed of sound in gas a long way from the black hole .
the last factor in ( [ bondac ] ) , q ( ) = ( 5 - 32)^3 - 5 2(-1 ) , is finite for both @xmath135 , when @xmath136 , and @xmath137 , when @xmath138 .
now consider a cooling flow .
thermal instability causes an inhomogeneous cooling flow to deposit gas throughout the flow at a rate that is conveniently expressed as ( nulsen 1986 , 1988 ; values there assume @xmath118 ) , where @xmath139 is the power radiated per unit volume by the gas , @xmath3 is now the mean density of the gas and @xmath140 is dimensionless .
the value of @xmath140 depends on details of the density distribution and physical behaviour of the inhomogeneous gas , and is typically of order unity .
we treat it as a constant parameter in what follows , since this leads to the simplest physically useful flow models ( this approximation is exact for the isothermal cooling flow models ; nulsen 1998 ) .
the mass conservation equation for a steady flow is then v r^2 = .
the corresponding energy equation is ( nulsen 1988 ) p v ddr = ( 1 - ) r , where @xmath141 , with @xmath20 being the temperature corresponding to the mean density ( @xmath3 ) at pressure @xmath119 .
@xmath142 determines the effective entropy of the inhomogeneous gas mixture . eliminating @xmath139 between these two equations and integrating gives = ( m ) ^(1-)/ , where @xmath143 and @xmath144 are evaluated at @xmath145 , a fixed point in the cooling flow , which we will take to be at the outer edge of the steady flow , well outside the region that is perturbed by the black hole .
thus , the cooling flow model enforces a fixed relationship between the entropy and @xmath146 throughout the steady flow .
the derivation of equation ( [ sigmdot ] ) makes no reference to the potential ( or the momentum equation ) and it applies at small @xmath8 , in the bondi flow where cooling is negligible .
the quantities @xmath128 and @xmath127 ( @xmath147 ) in the bondi accretion rate ( [ bondac ] ) are no longer well defined , but , since it is constant , the entropy , @xmath148 , is .
furthermore , ( [ sigmdot ] ) requires = ( ) ^(1-)/ , giving _
0 = ( t_0 ) ^1 -1 ( ) ^- ( 1- ) ( -1 ) .
we use this to eliminate @xmath128 in ( [ bondac ] ) , then solve the resulting equation for @xmath149 and put @xmath150 , where @xmath39 and @xmath33 are the temperature and density at @xmath145 . after some algebra
this gives = ^^1- ( t_0)^(5 - 3 ) 2 ( -1 ) , where @xmath151 .
this argument makes no reference to the gravitational potential in which the cooling flow takes place , but @xmath149 depends on details of the potential in two ways .
first , gas properties at the edge of the cooling flow are affected to some extent by the potential ( @xmath143 is governed largely by initial conditions in a collapse ) .
generally , we can assume that any disturbance to the potential is at @xmath152 and has little effect on the cooling flow at @xmath145 .
the second means of influence is through @xmath127 , which is still not well defined . despite this
, we can expect @xmath127 to be comparable to the gas temperature at the point where the black hole starts to have an appreciable affect on the cooling flow . since the influence of the black hole is felt outside the sonic point , this temperature will be comparable to the `` virial '' temperature at that place and , unless the galaxy potential is strongly non - isothermal , we can expect it to be comparable to @xmath39 . in general , the influence of the potential on @xmath149 is weak .
for the case of interest , @xmath32 and @xmath149 does not depend on @xmath127 , but is determined completely by the mass of the black hole and the gas properties near the outer edge of the cooling flow .
this simple result comes about because , for @xmath118 , the bondi accretion rate ( [ bondac ] ) depends on the gas properties through the entropy alone ( @xmath153 ) , so that it may be regarded as specifying a relationship between @xmath142 and @xmath146 .
the requirement ( [ sigmdot ] ) of the cooling flow specifies a second relationship between @xmath142 and @xmath146 which is only satisfied simultaneously by a unique @xmath146 . for an isothermal cooling flow , nulsen ( 1998 )
has shown that = q 3 - 2 ( ^2 ) k ( ) , where @xmath155 the flow speed at @xmath145 and @xmath156 is constant for the temperatures of interest .
the factor @xmath157 is a constant that depends on details of the cooling function . for power laws , @xmath25 , with @xmath26 in the range @xmath158 $ ] , @xmath157 ranges from @xmath159 to @xmath160 .
since @xmath161 lies in about this range for the temperatures of interest , we use the representative value @xmath162 in all models . using the expression for @xmath33 to eliminate the density in @xmath163 and using the result to replace @xmath143 in ( [ invmdot ] ) gives , after some further algebra , = ( 3- ) q kgm ^3 2 ( ) ^2 ( 2 ^2 ^12 gm)^3(1- ) 3 + , where @xmath164 is the mach number at @xmath145 .
observations show that @xmath165 in the well studied cluster cooling flows ( fabian 1994 ; peres 1998 ) and this is the value that we use for our models , giving = 2 q kgm ^3/2 ( ) ( ^2 ) .
we note that @xmath166 is not well determined and , since the factor in parentheses in ( [ penul ] ) is usually large , a small change in @xmath166 can have a large effect on @xmath149 . for the isothermal cooling flow model , the radial dependence of the mach number is @xmath167 , so that @xmath168 is constant in our models . if @xmath169 ( @xmath170 ) , then @xmath168 increases ( decreases ) inward .
in general , thermal instability in a cooling flow is greater when the mach number is larger ( balbus & soker 1989 ) , which will tend to limit the rise in the mach number for @xmath31 .
this limits the error in @xmath149 in that case . for @xmath170
there is no limiting effect , so that , if @xmath171 , ( [ mainres ] ) could substantially overestimate the nuclear accretion rate .
if so , the rate of accretion of hot gas would be insufficient to account for quasars . | we consider a model for quasar formation in which massive black holes are formed and fuelled largely by the accretion of hot gas during the process of galaxy formation . in standard hierarchical collapse models ,
objects about the size of normal galaxies and larger form a dense hot atmosphere when they collapse .
we show that if such an atmosphere forms a nearly `` maximal '' cooling flow , then a central black hole can accrete at close to its eddington limit .
this leads to exponential growth of a seed black hole , resulting in a quasar in some cases . in this model ,
the first quasars form soon after the first collapses to produce hot gas .
the hot gas is depleted as time progresses , mostly by cooling , so that the accretion rate eventually falls below the threshold for advection - dominated accretion , at which stage radiative efficiency plummets and any quasar turns off . a simple implementation of this model , incorporated into a semi - analytical model for galaxy formation , over - produces quasars when compared with observed luminosity functions , but is consistent with models of the x - ray background which indicate that most accretion is obscured .
it produces few quasars at high redshift due to the lack of time needed to grow massive black holes .
quasar fuelling by hot gas provides a minimum level , sufficient to power most quasars at redshifts between one and two , to which other sources of fuel can be added .
the results are sensitive to feedback effects , such as might be due to radio jets and other outflows . #
1to 0pt#1 # 1 = cmr10 scaled2 = cmbx10 scaled2 = cmti10 scaled2 ps .
# 1@xmath0 = 6truein = = = = = = = = # 1 # 1 # 1 # 1 @mathgroup@group @mathgroup@normal@groupeurmn @mathgroup@bold@groupeurbn @mathgroup@group @mathgroup@normal@groupmsamn @mathgroup@bold@groupmsamn = `` 019 = ' ' 016 = `` 040 = ' ' 336 = " 33e = = = = = = = = # 1 # 1 # 1 # 1 = = = = = = = = # 1 [ firstpage ] galaxies : formation quasars : general |
diffusion in heterogeneous and porous materials @xcite , composed of two or more types of regions with very different diffusion properties , can be described as a sequence of independent scattering events occurring in the hard - scattering part of the material , followed by long jumps performed at almost constant velocity in non - scattering regions .
the single - particle dynamics thus amount to a random walk , where each step length @xmath0 is a random variable with a given probability distribution @xmath1 .
as it is well known , if @xmath0 has a finite variance , the ensuing process follows the standard laws of brownian motion .
if instead the material is very heterogeneous on all scales , the step length distribution may become heavy - tailed and when @xmath2 diverges the diffusion can become anomalous @xcite . whenever the detailed structure of the underlying scattering media can be ignored ( _ annealed _ disorder ) , the dynamics is only ruled by the behavior of @xmath1 for large @xmath0 and the model corresponds to a standard lvy walk @xcite .
if instead one takes into account that the steps are correlated by their mutual positions in the sample , the step length distribution represents a _ quenched _ disorder .
it is argued that quenched and annealed disorder differ in many respects @xcite .
a particularly interesting application is multiple scattering of light in disordered media , which can often be described as a random walk process , analogous to the brownian motion of massive particles @xcite . in that case , if scattering elements are homogeneously distributed in space , one usually observes diffusive transport . however , when the local density of scattering elements is strongly inhomogeneous , the optical transport properties of a material can change dramatically and lead to superdiffusion , as exemplified for instance by photon diffusion through clouds @xcite .
more recently , artificial materials have been assembled in the lab that allowed for an unprecedented experimental demonstration of light superdiffusion @xcite .
interestingly , in experiments a change in the optical density of the diffusive media allows to tune the so called scattering length ( or scattering mean free path ) @xcite , which is a measure of the probability of experiencing a scattering event @xcite .
also monte carlo simulations @xcite indicate that this is an important parameter , that may significantly affect the observability of asymptotic scaling regimes .
it is thus important to asses the role of the scattering lengths on scaling properties of relevant observables , starting with probability distributions , to interpret correctly experiments and simulations in inhomogeneous scattering media .
while in pure diffusive processes a varying scattering length can be simply encoded in a trivial change of the variance of the gaussian distributions , superdiffusive process have not been investigated in details .
many relevant features of experiments on scattering in inhomogeneous media can be described as a random walk in a quenched , long - range correlated environment .
the simplest case consists of a free particle moving through a one - dimensional array of barriers whose spacing is power - law distributed @xcite . to address the question discussed above , in this work we extend such models to the case in which the transmittance through each barrier can be tuned to be different from @xmath3 .
this models a situation in which the velocity in the scattering media is not fully randomized at each collision . changing the barrier transmittance
is thus akin to changing the scattering length in the diffusive portion of the material . in the following
, we will show that the probability distribution for the random walker to be a time @xmath4 in the point @xmath5 exhibits a scaling form and that the effect of varying the transmittance can be reabsorbed in the multiplicative coefficient of the scaling length of the process .
numerical simulations confirm this superscaling behavior and evidence that not only the dynamical exponents are universal , as expected from the general scaling framework , but also the scaling functions are unchanged .
notice that the latter are in general not gaussian in these systems .
moreover , we obtain an analytic expression for the multiplicative coefficient of the scaling length as a function of the transmittance both in the annealed and in the quenched models ; in the annealed case the analytic form is calculated by solving directly the master equation of the process ; in the quenched models it can be derived , within the scaling hypothesis , by considering an extended einstein relation between the stationary conduction and the ( super)diffusion .
interestingly such coefficient turns out to be different in the quenche and in the annealed models and , in the latter case , it is independent of the transmittance when superdiffusion is present .
this evidences another relevant difference between the two approaches .
finally , we apply the same scaling picture , as a function of the transmittance , to the time resolved transmitted intensity through a finite sample of length @xmath6 and to the total transmission , which are the quantities usually measured in experiments .
the paper is organized as follows : in the next section we define the random and fractal quasi - lattices models with arbitrary transmittance , introducing the scaling picture and the superuniversal behavior . in section [ sec : annealed ] we calculate the exact expression , as a function of the transmittance , of the multiplicative coefficient for the scaling length in the annealed case . in section [ sec : quenched ] we turn to the quenched random and fractal quasi - lattices and we obtain an exact expression for the coefficients in these cases . in section
[ sec : numerics ] we analyze the superscaling framework for random structures with lvy parameter @xmath7 , a case with non gaussian scaling functions .
then , we focus on random structures with @xmath8 , where the gaussian scaling function obeys superscaling , but a subleading term determines the moments of the distribution .
we evidence that also this subleading term can be studied in terms of a change in the multiplicative constant for the scaling length .
we finally verify the superscaling framework in the deterministic case with @xmath9 where the scaling function presents log - periodic oscillations @xcite . in section [ sec : timeres ]
we apply our superscaling approach to a different physical quantity , the time resolved transmission intensity , which is relevant for a realistic experimental setup in inhomogeneous optical media . in the last section we present our conclusions .
we consider a continuous time random walk on one - dimensional structures ( see fig .
[ model ] ) , where point - like scatterers are spaced according to a lvy distribution .
the walker moves with constant velocity @xmath10 in between each two consecutive scatterers .
when it arrives at a scatterer , it can reverse its direction of motion ( @xmath11 ) with probability @xmath12 ( with @xmath13 ) .
clearly , @xmath14 is conserved during the evolution and we can set @xmath15 without loss of generality .
we investigate two types of structures .
the first is random and ( upper panel of fig .
[ model ] ) the probability for two consecutive scatterers , labeled by the indexes @xmath16 and @xmath17 , to be at distance @xmath5 is @xcite @xmath18 where @xmath19 and @xmath20 is a cutoff fixing the scale length of the system .
the second type is a class of deterministic quasi - lattices ( lower panel of fig .
[ model ] ) , built by placing the scatterers on generalized cantor sets @xcite .
each set , and the ensuing step length distribution , is defined by the two parameters @xmath21 and @xmath22 used in its recursive construction .
the former represents the growth of the longest step when the structures is increased by a generation , so that the longest step in a structure of generation @xmath23 is proportional to @xmath24 ; @xmath22 is the number of copies of generation @xmath25 that form the generation @xmath23 , so that the total number of scatterers in the generation @xmath23 is proportional to @xmath26 ( see ref.@xcite for details ) . for this second type of structures , the role of the exponent @xmath27 of the random case
is played by @xmath28 @xcite . in the random case
, we will average over different realizations of the structure and we will consider averages taken over processes starting from scattering sites .
for quenched lvy processes , it is known that different averaging procedures can lead to different behaviors .
moreover , properties arising from averages taken over processes starting in any point are different @xcite . in the deterministic case ,
we consider averages performed over random trajectories starting from a given point , e.g the origin @xmath29 evidenced in fig .
[ model ] .
the main quantity we are interested in is the probability for a walker to be at time @xmath4 a distance @xmath5 from the starting point , which we denote by @xmath30 to emphasize the dependence on the two basic parameters of this class of models , @xmath27 and @xmath31 .
the process on a lvy structure with given @xmath27 can be described by introducing a scaling function @xmath32 and a scaling length @xmath33 growing with time @xcite .
the scaling form for @xmath30 can be written as : @xmath34 where @xmath35 is a function that vanishes in probability for large times : @xmath36 in @xcite , the scaling form , , has been analyzed in details and it has been shown to hold for @xmath37 . moreover ,
the growth of @xmath38 has been shown to follow the asymptotic law @xmath39 for @xmath9 and @xmath40 for @xmath41 .
the presence of the subleading function @xmath42 and of long tails in @xmath32 , induce a nontrivial scaling of higher - order momenta , @xmath43 , leading to strongly anomalous diffusion @xcite . here
we will evidence that , within a superscaling framework , the dynamical exponents and the scaling functions are independent of the transmittance @xmath31 . in particular , @xmath44 is independent of @xmath31 and the scaling length reads : @xmath45 notice that @xmath33 is defined up to an arbitrary multiplicative constant , that in numerical simulation will be fixed equal to one for @xmath37 . within the hypothesis that @xmath44 is independent of @xmath31 , in the next sections we will obtain an analytic expression for @xmath46 both in the annealed and in the quenched case .
is transmitted with probability @xmath47 and reflected with probability @xmath12 ; ( b ) a deterministic cantor like lvy quasi - lattice , characterized by @xmath48 and @xmath49 . ]
before discussing the effect of the transmittance @xmath31 in quenched systems , we consider the annealed case , where the length of the ballistic stretches is chosen randomly from the distribution independently at each scattering event . in this case
the topology of the system is not taken into account and only the distribution of the steps influences the dynamics . in this simpler situation ,
an analytical approach is feasible and the behavior of the scaling length as a function of the transmittance can be determined .
let us introduce @xmath50 and @xmath51 as the probabilities of being in @xmath5 at time @xmath4 arriving respectively from the left and from the right . in the persistent random walk approach
, recalling that @xmath15 , we can write @xmath52 \lambda(r ' ) dr ' + \frac 1 2 \delta(r,0 ) \delta ( t,0)\nonumber \\ & & p^-(r , t ) = \int_{r_0}^\infty \left [ \frac { 1+\varepsilon } 2 p^-(r+r',t - r')+ \right .
\nonumber \\ & & + \left . \frac { 1-\varepsilon } 2 p^+(r+r',t - r ' ) \right ] \lambda(r ' ) dr ' + \frac 1 2 \delta(r,0 ) \delta ( t,0).\nonumber\\ \label{qmaster}\end{aligned}\ ] ] by applying a fourier transform in space and time in , we obtain an expression for @xmath53 , that is the transformed of @xmath54 .
in particular in the asymptotic regime for large space and times , i.e. for @xmath55 and @xmath56 going to zero , we have the following estimates : @xmath57 where the complex constants @xmath58 depend on the distribution @xmath59 ( i.e. on @xmath27 ) but are independent of @xmath31 .
the asymptotic scaling form of @xmath30 can then be obtained by inverting the fourier transform in . in particular we have : @xmath60 with @xmath61 where the function @xmath62 and proportionality constants in are independent of @xmath31 .
as expected , the dynamical exponents in equation coincides with the well known results for the annealed lvy walks with @xmath37 , established in @xcite .
in addition , equations ( [ panscal],[ell_an ] ) show that both the dynamical exponents and the scaling function @xmath62 are independent of @xmath31 , even in the non trivial case of anomalous diffusion i.e. @xmath63 . moreover , quite surprisingly for @xmath63 , i.e. in the ballistic and superdiffusive cases , even the coefficient of the scaling length does not depend on the transmittance and therefore the whole asymptotic regime is independent of the value of @xmath31 . on the other hand , for @xmath64
the same coefficient depends in a non trivial way both on @xmath31 and on the step length distribution @xmath59 .
indeed , @xmath65 where @xmath66 .
in particular , for the step lengh distribution considered in , @xmath67 . notice that the diffusivity diverges for @xmath68 ( perfect transmission ) , while for @xmath69 it vanishes only for @xmath70 , hence only when the step lengths do not fluctuate .
indeed , step length fluctuations induce diffusion even in the case of total reflection .
[ a_epsilon_an ] compares the analytical prediction for the coefficient in equations with numerical simulations at different value of @xmath27 showing an excellent agreement . .
symbols represent numerical simulations , while continuos lines are the analytical calculations . ]
let us now turn to the quenched case . within the scaling framework described by equations ( [ ellcom],[sal ] ) , we first derive an analytic expression for the coefficients @xmath71 .
we can exploit the fluctuation - response relation connecting @xmath30 to @xmath72 i.e. the stationary conductivity of a system of size @xmath6 .
in particular according to @xcite we have : @xmath73 plugging the scaling form of @xmath30 into equation and imposing that the scaling function depends on the transmittance @xmath31 only through the constant @xmath71 , as in , we obtain : @xmath74 where the proportionality constant is independent of @xmath31 .
equation extends the einstein relation for anomalous conductivity in @xcite , taking into account the transmittance @xmath31 .
the conductivity can be evaluated directly by studying the stationary current in a system of size @xmath6 and fixed boundary conditions .
the ( stationary ) master equation reads : @xmath75 where @xmath76 and @xmath77 represent the stationary probabilities of being at the scattering site @xmath78 arriving from the left and from the right , respectively .
the solution of equation is @xmath79 where @xmath80 and @xmath81 are arbitrary constants . imposing the particle density at the borders
@xmath82 and @xmath83 we get @xmath84 and @xmath85 where @xmath86 is the number of scatterers between @xmath29 and @xmath6 . since the conductivity equals the total current flowing in the system we have @xmath87 and averaging over different disorder realizations according to @xcite we have @xmath88 where the constant @xmath89 is independent of @xmath31 .
the results in equations is consistent with and , with an analytical estimate for the multiplicative coefficients of the scaling lenghts : @xmath90 in fig .
[ a_epsilon ] we compare our analytical prediction with numerical simulations at different value of @xmath27 , with excellent agreement .
clearly @xmath71 diverges for @xmath91 , i.e. perfect transmission and @xmath71 vanishes for @xmath92 , i.e. total reflection .
we remark that @xmath71 behaves differently than in the annealed model both in the superdiffusive and in the normal case . as a function of @xmath31 .
symbols represent numerical simulations , while continuos lines are the analytical calculations of formula . ]
let us now present a numerical analysis of the quenched cases , evidencing that our assumption holds , i. e. that the whole effect of a variation in @xmath31 can be summarized in a variation of the coefficient @xmath71 , and therefore that the scaling functions are super - universal , i.e. @xmath32 in eq . does not depend on @xmath31
. for standard diffusion this properties obviously holds : indeed the scaling functions are gaussian and they can be characterized basically by their variance , i.e. the scaling length of the process . however in superdiffusive processes , for @xmath93 , @xmath32 is a non trivial function decaying at large distances as @xmath94 , so an analogous property is not trivial . numerical data , in fig .
[ scal0_7 ] evidence for @xmath95 in the random case that @xmath30 for different values of the transmittance can be scaled into a single function , independent of @xmath31 . moreover
, the dashed line shows that the scaling function is different from the standard lvy function describing the sum of independent lvy distributed random variables , even if they are characterized by the same long tail @xcite . for @xmath95 on a random structure .
the value of @xmath33 are evaluated according to and the values of the coefficients @xmath71 is given by .
red - dashed line represents the shape of the scaling function characterizing an uncorrelated lvy flight with the same @xmath95 .
notice that the tail of the distribution is the same since in both processes it is determined by the value of @xmath27 . ]
let us now consider the random case with @xmath41 .
the scaling length grows as in a diffusive process , and therefore we expect the scaling function @xmath32 to be a gaussian , independently of @xmath31 and @xmath27
[ scal1_5 ] , obtained for @xmath96 , evidences this property for small @xmath97 .
however , for large @xmath97 , @xmath30 contains now the subleading contribution @xmath35 .
this term vanishes for large times but it can influence the momenta of the distribution . fig .
[ scal1_5 ] shows , at large @xmath97 , the presence of this subleading term . in @xcite , @xmath42 has been evaluated using a `` single long jump '' approximation . here
, we show that the same argument applies within the ansatz with the coefficients determined by , and leads to a correct estimate of the function @xmath35 . in particular , within the same approximation , for @xmath98 the probability of reaching a point at distance @xmath5
is determined by the probability of performing a single ballistic stretch of length @xmath5 , times the number of scatterers visited by the walker in a time @xmath4 ; this number can be estimated as @xmath99 , where @xmath100 is the average distance between the scatterers , that for @xmath41 is finite and independent of @xmath31 . if the whole effect of a variation in the transmittance @xmath31 can be encoded in a change of @xmath71 as in eq .
, then we expect that the behavior of @xmath35 can be estimated as @xmath101 where the proportionality constant is independent of @xmath31 . in fig .
[ scal1_5long ] we check indeed that our ansatz is correct and that the tail of the distribution @xmath30 can be described according to equation , with the @xmath31-dependent coefficient . for @xmath96 on a random structure .
the value of @xmath33 are evaluated according to and the values of the coefficients @xmath71 is given by .
dashed line represents the gaussian behavior of the scaling function for not too large @xmath97 . at larger @xmath97
there is a slowly decaying function @xmath35 superimposed to the gaussian behavior . at long times ( black symbols )
@xmath35 tends to vanishes , however it still influences the momenta of the distribution . ] as a function of @xmath5 for @xmath96 on a random structure .
the value of @xmath33 are evaluated according to equation .
the plot evidences that the tail @xmath35 of the distribution corresponds to equation where the effect of different transmission coefficient can be summarized in the constant @xmath71 in . ]
finally , we focus on the case of deterministic one - dimensional fractal quasi - lattices i.e. the lower panel of fig .
[ model ] , where the step length distribution is described by the parameters @xmath21 and @xmath22 . as explained in @xcite , for @xmath102 the motion of the random walker
is ruled by the parameter @xmath103 which plays the role of @xmath27 in the random structure . in particular , the scaling length of the process also in the deterministic case grows as @xmath104 for @xmath93 and @xmath105 for @xmath41 . here ,
we consider averages performed over processes starting from a given point of the structure , e.g the origin @xmath29 evidenced in fig .
[ model ] . for local quantities ,
the scaling function does not present long tails since arbitrary long jumps are placed far away from the starting point . on the other hand ,
for @xmath93 the fractality of the structure induces characteristic log - periodic oscillations in the scaling function . in particular for @xmath37 we have @xcite @xmath106 where @xmath38 is the scaling length of the process on the quasi - lattice , @xmath107 is the scaling function and @xmath108 is a function of period one . according to eq .
, a variation of the transmittance only induces a rescaling of the correlation length , so that the scaling function for a generic @xmath31 is expected to be : @xmath109 where @xmath33 and the corresponding coefficients are again given by equations ( [ ellcom],[aepsilon ] ) .
therefore for processes with different transmittances @xmath31 and @xmath110 , scaling holds if times @xmath4 and @xmath111 are chosen so that @xmath112 with @xmath56 integer , i.e. @xmath113 for @xmath114 on a fractal quasi - lattice
. the value of @xmath33 are evaluated according to ( [ ellcom],[aepsilon ] ) .
times are chosen according equation to detect scaling . ] in fig .
[ scallcantor ] we evidence that for times chosen according to equation scaling holds for different @xmath31 .
the complex devil staircase shape of the scaling function is typical of these fractal structures .
notice that the choice of times , as a function of @xmath31 , is crucial to recover the scaling , a further test of the validity of eq . .
in the case @xmath41 the scaling functions are gaussians , log - periodic oscillation are absent and scaling can be recovered through and , without the tuning of times , as in the random case .
due to its generality , the scaling approach can be useful not only in the analysis of @xmath30 but also for other interesting physical quantities , such as the exit time probability and the time resolved intensity @xcite , which are experimentally more relevant .
let us consider the effect of a transmittance coefficient @xmath115 in the case of a random structure .
we consider a walker starting at time @xmath116 from the border of a sample of size @xmath6 and we define the time resolved transmitted intensity @xmath117 as the probability for the walker to reach the boundary at distance @xmath6 before returning to the starting point .
the scaling hypothesis for the intensity @xmath117 reads @xmath118 where @xmath119 is a scaling function and
the proportionality constant @xmath120 depends on @xmath6 , @xmath27 and @xmath31 .
we calculate the coefficient @xmath120 as follows .
the conductivity @xmath72 is by definition the total number of walkers escaping from a system of size @xmath6 before returning to the starting point , independently of time i.e. @xmath121 imposing the integral to be proportional to @xmath122 according to equation , we obtain @xmath123 the scaling form has been verified in figure [ transm ] for @xmath124 and different values of @xmath31 , with the values of the coefficients determined by . .
we have shown that lvy walk in a quenched , long - range correlated structures satisfies a generalized scaling relation for arbitrary values of the transmittance , both in the random and in the fractal case , as expressed by eq .
( [ sal ] ) and ( [ sal_cant_ep ] ) respectively .
the main difference between the two models is in the presence of the subleading term for the random case and of log - periodic oscillations in fractal quasi - lattices .
as expected , all the leading scaling behavior are unaffected by a change of the scattering length ( the parameter @xmath31 in our model ) .
this parameter enters in the multiplicative prefactor , ruling the dependence of the scaling length as a function of time .
estimation of the latter is of course relevant for finite samples and times .
we obtained an analytic expression for the multiplicative coefficients of the scaling lengths , as a function of @xmath31 in the annealed lvy walk case , evidencing its independence on the transmittance in the superdiffusive regimes . within the scaling framework
, we also determined a closed form for the coefficient in the quenched random and fractal cases , which are the most relevant for experiments .
another remarkable result of our analysis is that the scaling functions feature a superscaling property , namely they are independent of the transmittance and only depend on the structure through the exponent @xmath27 .
interestingly , we employed the scaling properties to infer the dependence of the time - resolved transmission in finite samples .
all our analytic results have been compared with numerical simulations , with excellent agreement .
as the scaling picture discussed here has been shown to hold also in higher dimensional cases @xcite , we expect that this superuniversality , holding in one dimension , can be detected also in higher dimensions .
moreover , we expect that superuniversality could hold not only for a variation of the transmittance @xmath31 but for a wider class of local transformation of the dynamics such as the introduction of waiting times or of second neighbors jumps .
besides their theoretical interest , our result are of importance to interpret correctly the experimental and numerical results .
for instance , in the optical experiments of ref .
@xcite it is possible to control the mean free path of light in diffusive media and investigate the approach to the scaling limits for the same distribution of glass sphere diameters .
this work has been partially supported by the miur project _ p.r.i.n .
2008 _ `` nonlinearity and disorder in classical and quantum processes . '' and by the miur project _ p.r.i.n .
2008 _ project `` efficienza delle macchine termoelettriche : un approccio microscopico '' .
the authors acknowledge useful discussions with p. buonsante .
fogedby , phys .
73 * 2517 ( 1994 ) , r. kutner and p. maass , j. phys .
a : math . gen . * 31 * , 2603 ( 1998 ) , m. schulz , phys .
a , * 298 * , 105 ( 2002 ) .
e. barkai , v. fleurov , j. klafter , phys .
e * 61 * 1164 ( 2000 ) .
groth , a.r .
akhmerov , c.w.j .
beenakker , phys .
e * 85 * , 021138 ( 2012 ) . | we study the effects of scattering lengths on lvy walks in quenched one - dimensional random and fractal quasi - lattices , with scatterers spaced according to a long - tailed distribution . by analyzing the scaling properties of the random - walk probability distribution ,
we show that the effect of the varying scattering length can be reabsorbed in the multiplicative coefficient of the scaling length .
this leads to a superscaling behavior , where the dynamical exponents and also the scaling functions do not depend on the value of the scattering length . within the scaling framework
, we obtain an exact expression for the multiplicative coefficient as a function of the scattering length both in the annealed and in the quenched random and fractal cases .
our analytic results are compared with numerical simulations , with excellent agreement , and are supposed to hold also in higher dimensions . |
in the early days of pulsar observations , slow fluctuations of pulsar amplitudes were reported by @xcite and @xcite .
but it was not recognized that they were an interstellar propagation effect until a close correlation between the time - scale of the intensity fluctuations and the pulsar dispersion measure was pointed out by @xcite .
subsequently , @xcite explained these slow fluctuations as refractive interstellar scintillations ( riss ) .
this phenomenon is produced by the same spectrum of electron density fluctuations in the interstellar medium that is responsible for the well known diffractive scintillation ( diss ) .
diss is caused by the small spatial scale density fluctuations ( @xmath2 m ) and it appears as intensity variations in both the time and frequency domains with characteristic scales of minutes to hours and khz to mhz respectively .
on the other hand , large - scale inhomogeneities ( @xmath3 m ) in the interstellar medium give rise to focussing effects observed as riss .
a good review of the theory behind this interpretation is given by @xcite .
it is generally accepted that the spectrum of electron density fluctuations in the interstellar medium has a power - law form : @xmath4 where @xmath5 is the wave - vector associated with a spatial size @xmath6 .
the quantity @xmath7 is a measure of turbulence along a particular line of sight .
the power - law index @xmath8 which indicates the steepness of the inhomogeneity spectrum is in the range of @xmath9 .
but the details of the spectrum , e.g. its slope and the range of scale sizes over which it is valid , still remain open questions .
also , different assumptions are made about the distribution of scattering material along the path to the pulsar .
many observations are well accounted for by models in which the scattering is concentrated in a single centrally located thin screen , but others imply that there are many scattering sites or a statistically uniform screen extending over most or all of the path .
a range of observations suggest that a kolmogorov fluctuation spectrum with @xmath10 is widely applicable in the interstellar medium @xcite . however , observational studies showing , for example , persistent drifting bands in dynamic spectra and increased modulation of diffractive properties suggest that excess power is often seen at low spatial frequencies , implying a spectral index @xmath11
. examples of extensive studies are those by @xcite , who studied eight pulsars at 408 mhz using the lovell telescope and interpreted changing diffractive patterns in dynamic spectra in terms of refractive effects , and those by @xcite and @xcite .
these latter authors used the ooty radio telescope at 327 mhz to study 18 pulsars over a 2.5-year dataspan and showed that , while many observations were consistent with a kolmogorov spectrum , others showed evidence of a steeper spectrum .
@xcite reported on 5 years of daily monitoring of the flux density of 21 pulsars at 610 mhz , showing that most of the results , covering a wide range of scattering strengths , were consistent with a kolmogorov fluctuation spectrum .
in some pulsars , though modulation indices were higher than expected , possibly due to the presence of an `` inner scale '' , a cutoff in the fluctuation spectrum at scales greater than the diffractive scale @xmath12 , where @xmath13 is the wavelength and @xmath14 is the angular size of the scattering disk . other observations suggesting steeper power - law spectra ( @xmath15 ) include fringing in dynamic spectra implying multiple imaging ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and extreme scattering events @xcite .
@xcite suggested a steeper spectrum to deal with the theoretical difficulties in supporting a turbulent cascade in the kolmogorov spectrum .
based on power - law models of the spectrum , @xcite analysed the effects of riss on diffractive parameters for different slopes @xmath10 , 4 and 4.3 , predicting anti - correlations for ( @xmath16,@xmath17 ) and ( @xmath16 , @xmath18 ) , where @xmath16 is flux density and @xmath17 , @xmath19 are the diffractive scintillation bandwidth and timescale , respectively and a postive correlation for ( @xmath17,@xmath19 ) , with the magnitudes of the correlation coefficients being larger for steeper spectra .
more recent observations have tended to concentrate on detailed studies of individual pulsars ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) or on the fascinating `` scintillation arcs '' which are seen in secondary spectra , that is , two - dimensional fourier transforms of dynamic spectra , in high - sensitivity , high - resolution observations @xcite .
these arcs , which are closely related to multiple imaging and the frequently observed `` criss - cross '' sloping bands , result from interference between rays in a central core and rays from an extended scattering disk .
they form a powerful probe of structure in the interstellar medium @xcite .
for example , @xcite present recent observations of six pulsars showing multiple arcs of different curvature implying a distribution of scattering centres along the line of sight .
psr b0329 + 54 is one of the strongest pulsars known and its scintillation properties have been studied by many authors .
observations at 610 mhz by @xcite showed that the correlations between variations of flux , decorrelation bandwidth and scintillation time - scale were consistent with the theoretical predictions with @xmath20 .
however , @xcite found that for psr b0329 + 54 , correlations between these parameters at 327 mhz were not in accord with the predictions .
long - term flux density monitoring at 610 mhz by @xcite showed that the modulation index was somewhat higher than expected for a kolmogorov spectrum , i.e. @xmath21 .
@xcite took data obtained over a wide range of frequencies finding a value of @xmath22 , less than but consistent with a kolmogorov turbulence spectrum , and interpreted this result in terms of weak plasma turbulence in the interstellar medium .
observations at 1540 mhz using the nanshan telescope @xcite showed inconsistencies in values of the power - law index obtained from two different approaches .
the ratio of refractive scattering angle @xmath23 and diffractive angle @xmath24 gave @xmath25 , while the frequency dependence of scintillation parameters gave @xmath26 .
these conflicting results suggested undertaking observations with a better sampling of the refractive variations .
these have a predicted timescale @xmath27 @xcite , about four days for this pulsar . in this paper , we present the results of quasi - continuous 1540 mhz observations of psr b0329 + 54 over a 20-day interval in 2004 , march .
diffractive parameters were sampled at 90-min intervals , giving good resolution of the expected riss variations .
the layout for the rest of this paper is as follows : in section 2 we introduce the observations ; data analysis and results are shown in section 3 , and in section 4 , we discuss the results and their interpretation .
conclusions are presented in section 5 .
the observations were made using the nanshan 25 m radio telescope with a central frequency of 1540 mhz .
our system is sensitive to orthogonal linear polarizations and the system temperature for the two channels is 20 k and 22 k respectively .
each polarization has 128 channels , each of bandwidth 2.5 mhz , giving a total bandwidth of 320 mhz . after detection and high - pass filtering ,
the data are 1-bit sampled and folded at the apparent pulsar period .
signals from the two polarizations are summed to form total intensity pulse profiles .
we made our observations from march 12 to march 31 in 2004 , and obtained more than 150 scintillation dynamic spectra for psr b0329 + 54 .
the observations were continuous ( psr b0329 + 54 is circumpolar at nanshan ) apart from an interruption of about 3 days from march 20 due to telescope problems . in our analysis , the time block for each dynamic spectrum is typically 90 minutes with a sub - integration time of 60 s. we calibrated the pulsar flux density scale using observations of ten strong , relatively distant pulsars which all have well - measured flux densities in the atnf pulsar catalogue ( www.atnf.csiro.au/research/pulsar/psrcat ) .
dynamic spectra were obtained for each observation by plotting the mean flux density in each channel against time .
examples of these plots are shown in fig .
[ fg : dynamic ] .
channels badly affected by radio - frequency interference have been interpolated over from adjacent frequency channels . fig .
[ fg : dynamic ] shows that the dynamic spectra are well resolved and that the scintles or peaks of flux density change their shapes and sizes in the frequency and time domains . at some epochs , the pulsar seems weaker and the scintles are smaller ( mjd 53088.4 ) , and at other epochs , the pulsar signal is stronger and the scintles larger ( e.g. mjds 53081.6 , 53083.2 , 53093.3 ) . the frequency dependence of diss can be seen within our receiver bandwidth , i.e. , scintles are smaller at the lower end of the observed band ( mjd 53091.8 ) . at some epochs
the spectra show a significant drifting pattern ( e.g. mjd 53082.4 ) , which implies a modulation by riss .
however , we did nt observe the fine fringes seen in the data of @xcite .
the scintles are generally smaller in this observing session , resulting in a smaller value in @xmath28 , which we will discuss more in section [ sec : gauss ] .
deep modulations are common in all the dynamic spectra we have recorded . the limited number of scintles in the dynamic spectra introduces a statistical estimation error in the scintillation parameters @xcite . an approximate estimate of the number of scintles @xmath29 in the dynamic spectra is given by : @xmath30 where @xmath31 is the total observing time and @xmath32 the total observing bandwidth , @xmath28 decorrelation bandwidth and @xmath19 is the diss timescale .
the fractional estimation error is then given by : @xmath33 where we assume a scintle filling factor of 0.5 @xcite .
this error contribution is taken into account in the estimates of scintillation parameters in section [ sec : gauss ] .
the two - dimensional fourier spectrum of the primary dynamic spectrum is often referred to as the secondary spectrum . to show the beautiful arc structures ( e.g. , * ? ? ? * ) , high frequency resolution and high sensitivity are required .
unfortunately , the nanshan system does not provide these .
however , earlier observations at 1540 mhz for this pulsar by @xcite showed sloping fringes in the dynamic spectra resulting in offset features in the secondary spectra .
surprisingly , despite using the same system at nanshan , our three - week consecutive observations show no significant arc structure or offset features in the secondary spectra .
to quantify the diffractive parameters , we use the two - dimensional auto - correlation function ( acf ) , which was computed for frequency lags up to half of the observing bandwidth and for time lags up to half of the observing time . the acf of the dynamic spectra @xmath34 is defined as : @xmath35 where @xmath36 , and @xmath37 is the mean pulsar flux density over each observation , then the normalized acf is : @xmath38 we plot the normalized acf and one dimensional cut at zero time and frequency lag in fig .
[ fg : contour ] .
these figures correspond to the dynamic spectra shown in fig .
[ fg : dynamic ] .
following convention , the diss time - scale @xmath19 is defined as the time lag at zero frequency lag where the acf is 1/e of the maximum , and the decorrelation bandwidth @xmath28 is defined as the half - width at half - maximum of the acf along the frequency lag axis at zero time lag @xcite . for weaker pulsars , random system noise results in a strong spike at the zero time and frequency lag of acf . for psr b0329 + 54
the spike is weak but still observable at some epochs . to remove the spike at the origin we used the four neighbouring frequency - lag points to fit for a parabola across the zero frequency lag .
the central point is interpolated from this fit .
following @xcite and @xcite , we use a two - dimensional elliptical gaussian function to fit the acf with form : @xmath39,\ ] ] in which @xmath40 is unity since the acf is normalized to unity . by using a @xmath41 minimization procedure , we obtained parameters @xmath42 , @xmath43 and @xmath44 .
the fitting procedure was a search for the three non - linear parameters for the least - squared error between model and data over the central region of acf .
the scintillation parameters @xmath28 and @xmath19 are calculated as : @xmath45 following @xcite , we use @xmath46 to describe the orientation of the elliptical gaussian .
this is proportional to the refractive scattering angle @xmath47 , and is given by : @xmath48 . from the @xmath49 analysis
, we also obtained the uncertainties for @xmath42 , @xmath43 and @xmath44 . using these uncertainties added in quadrature with the statistical error ( equation [ eq : staerr ] )
, we calculate the errors of @xmath50 , @xmath19 and @xmath46 .
since @xmath46 has zero mean , we applied the statistical error to @xmath51 rather than to @xmath43 itself .
for flux density , measurement errors are estimated from the baseline noise in the mean pulse profiles .
time variations of @xmath16 , @xmath28 , @xmath52 and @xmath46 along with their uncertainties are shown in fig .
[ fg : para ] .
these parameters show significant variations with time which can be explained as an riss effect as discussed in section [ sec : interpretation ] . given a power - law spectrum of electron density fluctuations ( equation [ eq : spectrum ] ) , the average scattering level along the line - of - sight , @xmath7 , for a kolmogorov spectrum is given by @xmath53 @xcite .
another quantitative measurement of the strength of scattering is the parameter @xmath54 which is defined as the ratio of the fresnel scale @xmath55 to the coherence scale @xmath56 @xcite , where @xmath14 is the diffractive scattering angle . for strong scattering , @xmath57 . in terms of the diffractive scintillation parameter @xmath28 : @xmath58 the diffractive timescale depends on the velocity of the ray path across the scattering screen .
assuming a thin screen and that the pulsar velocity dominates , an estimate for the pulsar transverse velocity based on the scintillation parameters is : @xmath59 where @xmath60 is the ratio of the observer - screen distance to the pulsar - screen distance and the proportionality constant @xmath61 @xcite . in table
[ tb : param ] we give some basic parameters for the pulsar and observations as well as mean values for the basic scattering parameters .
the pulsar distance is based on a parallax measurement by @xcite who also measured the pulsar proper motion , @xmath62 mas yr@xmath63 which , with the distance , gives the transverse velocity listed in the table . the typical observation time for each dynamic spectrum and the total number of such observations are given .
lr galactic longitude ( deg ) & 145.00 + galactic latitude ( deg ) & @xmath64 + dm ( pc @xmath65 ) & 26.83 + distance ( kpc ) & @xmath66 + @xmath67 ( km s@xmath63 ) & @xmath68 + observation time ( min ) & 90 + + number of observations & 168 + @xmath69 ( mhz ) & @xmath70 + @xmath71 ( min ) & @xmath72 + @xmath73 ( @xmath74 ) & @xmath75 + @xmath76 ( h ) & @xmath77 + @xmath7 ( @xmath78 ) & @xmath79 + @xmath54 & @xmath80 + @xmath81 ( km s@xmath63 ) & @xmath82 + in the lower part of the table we give the mean values and rms scatter for the decorrelation bandwidth ( @xmath28 ) , scintillation time - scale ( @xmath19 ) and the slope parameter ( @xmath46 ) derived from the acf analysis .
we note that the scintillation timescale @xmath19 is much less than the observation time ( 90 min ) , so the effects of finite observation time discussed by @xcite should be minimal .
the predicted refractive timescale , @xmath76 based on the diffractive parameters ( equation [ eq : tref ] ) , is about 4 d , but the observed value ( section [ sec : ref - time ] ) is much less than that .
the derived scattering level , @xmath7 , and scattering strength , @xmath54 , both indicate that scattering is strong along the path to psr b0329 + 54 . for the quoted scintillation velocity ,
a centrally located screen ( @xmath83 ) has been assumed .
our value is somewhat lower than the proper motion velocity .
it is much lower than the value given for this pulsar by @xcite based on 408 mhz observations , but they assumed a pulsar distance of 2.3 kpc . for the parallax distance of @xmath84 kpc ,
their velocity becomes @xmath85 km s@xmath63 , consistent with the proper motion velocity . from their 327 mhz observations
@xcite derive a scintillation velocity of @xmath86 km s@xmath63 based on an assumed pulsar distance of 1.43 kpc .
scaled with the parallax distance , this becomes @xmath87 km s@xmath63 .
while these results are not too discordant , it is clear that there are variations on timescales much longer than the nominal refractive time and also that the frequency scaling of the diffractive parameters does not precisely follow the kolmogorov prediction ( cf .
@xcite suggested that correlations should exist between variations of the flux density @xmath16 , decorrelation bandwidth @xmath28 , and diffractive timescale @xmath88 .
these predictions assume a single - phase screen and a simple power - law description for the electron density fluctuations . in table
[ tb : cross ] we list the predicted correlation coefficients for turbulence spectra of @xmath8911/3 , 4 and 4.3 @xcite . [ cols="<,^,^,^,^",options="header " , ] another possible intepretation of the relatively large observed modulation indices is that the scattering medium is extended so that the thin - screen approximation is not valid .
the flux density modulation index for a statistically uniform scattering medium with a kolmogorov spectrum ( @xmath90 ) covering the whole path to the pulsar has been discussed by @xcite , @xcite , @xcite and @xcite with the result @xmath91 for our observations of psr b0329 + 54 , this relation gives a predicted @xmath92 , larger than the observed value .
this implies that , as an alternative to a steeper spectrum , the observed modulation index could be accounted for by an extended kolmogorov scattering medium covering just part of the path .
the horizontal dashed lines on the structure function plots ( fig .
[ fg : struct ] ) are at @xmath93 , where @xmath94 is the observed modulation index given in in table [ tb : mod ] .
the refractive timescale @xmath76 is defined to be the lag at which the structure function falls to @xmath95 . in the flux density case , this is at @xmath77 h. the structure functions for the diffractive bandwidth and timescale are less well defined , but clearly also imply a short refractive timescale .
the measured timescale is much less than the predicted value of about 96 h based on a central thin - screen model and the observed diffractive timescale ( table [ tb : param ] ) .
for an extended scattering medium with a kolmogorov spectrum , evaluating the equations of @xcite with practical units gives @xmath96 substituting the parameters from our observations of psr b0329 + 54 gives a value of about 32 hours , again much larger than the observed value .
@xcite also consider the case of a thin screen at an arbitrary location along the path and show that @xmath76 is proportional to the pulsar - screen distance ( @xmath97 ) whereas @xmath19 is proportional to @xmath98 .
therefore , if @xmath99 , refractive times are much reduced , whereas @xmath19 is only weakly dependent on @xmath97 .
this breaks the approximate proportionality of @xmath76 and @xmath19 ( equation [ eq : tref ] ) which applies to both a central thin screen and an extended medium .
the scintillation arc observations of @xcite suggest that most of the scattering for psr b0329 + 54 occurs relatively close to the pulsar , supporting this interpretation of the short refractive timescale .
the slope of the linear part of the structure function , between the noise and saturation regimes , is related to the power - law index of the fluctuation spectrum . for a single thin screen
, the slope should be 2.0 for @xmath100 @xcite .
for an extended screen , the predicted slope is @xmath101 @xcite . although the observed slopes are not very well determined , they are clearly @xmath102 . the measured value for the flux density structure function slope is @xmath1 and the values for the other two structure functions are similar .
this is clearly inconsistent with scattering by a single thin screen .
as we have argued above , there are good arguments supporting the idea that the scattering screen in this direction is extended ; in this case , the implied value of @xmath103 , somewhat less than but marginally consistent with the kolmogorov value . based on the thin - screen model ,
observed diffractive bandwidths and a refractive interpretation of band slopes ( @xmath46 ) , @xcite and @xcite also obtain @xmath8 values less than the kolmogorov value for psr b0329 + 54 and several other pulsars .
our result also agrees well with that from @xcite who obtain a structure function slope of @xmath104 from their flux density monitoring of this pulsar .
however , it contrasts with that of @xcite where a structure function slope @xmath105 is derived from observations at several different frequencies and interpreted in terms of weak plasma turbulence , again with a kolmogorov spectrum .
quasi - continuous observations of psr b0329 + 54 at 1540 mhz over 20 days have been analysed to investigate the refractive modulations of the pulsar flux density and diffractive scintillation parameters .
the data set was split into more than 150 individual observations , each of 90-min duration , and two - dimensional auto - correlations and secondary spectra computed .
time series of the pulsar flux density ( @xmath16 ) , decorrelation bandwidth ( @xmath28 ) , diffractive scintillation time - scale ( @xmath19 ) and drift rate of features in the dynamic spectra ( @xmath46 ) clearly show refractive modulation , although no evidence for arc structure is seen in the secondary spectra . observed cross - correlations between variations in the flux density and diffractive parameters were much smaller than predictions based on the thin - screen model .
in one case the sign of the correlation was opposite to the predicted value .
similar results have been obtained by other observers although different observations appear to give conflicting results . in accordance with previous work , observed modulation indices are greater than predicted for a thin screen with a kolmogorov fluctuation spectrum .
this could be accounted for either by a steeper fluctuation spectrum ( @xmath106 ) or by scattering in an extended medium .
the predicted modulation index for a scattering medium covering the whole path to the pulsar is greater than that observed , suggesting that the scattering medium , while extended , does not cover the whole path .
structure functions derived from the observations indicate a short refractive timescale , @xmath0 h , much less than predicted from the thin screen model , and have a relatively flat slope , @xmath1 , again inconsistent with scattering by a thin screen .
the observed modulation indices , structure function slopes and short refractive timescales all are consistent with scattering by an extended region with a kolmogorov fluctuation spectrum which is concentrated toward the pulsar .
this idea is supported by recent high - sensitivity observations of psr b0329 + 54 by @xcite which show indistinct scintillation arcs corresponding to extended scattering regions relatively close to the pulsar .
we would like to thank b. j. rickett and w. a. coles for helpful discussions .
this work is supported by the key directional project of cas and nnsfc under the project 10173020 and 10673021 . | quasi - continuous observations of psr b0329 + 54 over 20 days using the nanshan 25-m telescope at 1540 mhz have been used to study the effects of refractive scintillation on the pulsar flux density and diffractive scintillation properties .
dynamic spectra were obtained from datasets of 90 min duration and diffractive parameters derived from a two - dimensional auto - correlation analysis .
secondary spectra were also computed but these showed no significant evidence for arc structure .
cross correlations between variations in the derived parameters were much lower than predicted by thin screen models and in one case was of opposite sign to the prediction .
observed modulation indices were larger than predicted by thin screen models with a kolmogorov fluctuation spectrum .
structure functions were computed for the flux density , diffractive timescale and decorrelation bandwidth .
these indicated a refractive timescale of @xmath0 h , much shorter than predicted by the thin screen model .
the measured structure - function slope of @xmath1 is also inconsistent with scattering by a single thin screen for which a slope of 2.0 is expected .
all observations are consistent with scattering by an extended medium having a kolmogorov fluctuation spectrum which is concentrated towards the pulsar .
this interpretation is also consistent with recent observations of multiple diffuse scintillation arcs for this pulsar .
[ firstpage ] pulsars : individual : psr b0329 + 54 ism : structure |
the production of hadrons at large transverse momentum @xmath4 ( in hadron - hadron or hadron - photon collisions ) offers a classical test of perturbative qcd .
the cross sections of the hard subprocesses are calculated with next - to - leading - order ( nlo ) accuracy @xcite , the quark and gluon distributions in the initial hadrons are measured in deep - inelastic - scattering experiments @xcite and the fragmentation functions describing the transitions of the partons into the final - state hadrons are extracted from @xmath5-annihilation data @xcite .
therefore , all the building blocks of the large-@xmath4 cross sections are in principle known , and the comparison with data provides interesting tests of the theory
. the hadroproduction of large-@xmath4 @xmath0 mesons has been studied by the cggrw collaboration @xcite , who obtained good agreement between the qcd predictions and collider @xmath0 cross sections . on the other hand , fixed - target data with a lower center - of - mass
energy @xmath6 ( 23 gev@xmath7 gev ) overshoot the theoretical predictions , and it was impossible to obtain an overall agreement with all the large-@xmath4 @xmath0 data . large-@xmath4
charged - hadron cross sections were studied in hadroproduction at collider energies @xcite and in photoproduction at hera energies @xcite . in that case , reasonable agreement between theory and data was reached , but one must remember that these predictions are sensitive to the choice of the factorization and renormalization scales in the @xmath4-range which was studied ( @xmath8 gev ) . in this paper , we come back to the study of the large-@xmath4 @xmath0 cross sections . new fixed - target data appeared recently @xcite , completing those already available in the range @xmath9 gev @xcite . on the theoretical side , new sets of fragmentation
function extracted from lep data are now available @xcite .
one must also notice that @xmath10 has increased from @xmath11 200 mev to @xmath11 300 mev since the time of the first study @xcite .
this new value of @xmath12 gives rise to a non - negligible increase of the qcd cross section in the @xmath4 range studied here ( @xmath8 gev ) .
another important reason for this study is the publication of new data on prompt - photon cross sections @xcite .
a recent qcd analysis of all fixed - target and isr data @xcite leads to the conclusion that theory is in reasonable agreement with experiment , with the exception of two data sets , as shown in fig .
[ xtcteqpt2 ] .
however , it should be remembered that the comparison with qcd predictions can only be done for high - enough values of @xmath4 ( _ e.g. _ @xmath13 gev / c for the e706 and isr energy ranges ) , where the scale dependence of the theoretical cross sections is not too large .
therefore , in fig . [ xtcteqpt2 ] , one should not consider the first points of the e706 and isr experiments .
the introduction of an additional intrinsic transverse momentum @xmath14 of the incoming partons , strongly advocated in refs .
@xcite to enforce agreement between qcd predictions and experiment , is not a very satisfactory solution ; the agreement with the e706 data sets is improved , but the agreement with other data sets is reduced .
this situation is challenging and leads us to look at large-@xmath4 @xmath0 production .
the large-@xmath4 @xmath0 mesons form a significant background for the prompt photons , and their cross section must be carefully measured in order to allow for a reliable estimate of the `` fake prompt photons , '' in particular due to configurations in which one photon from the decaying @xmath0 escapes detection .
therefore , there is a strong experimental correlation between the prompt - photon and the @xmath0 cross sections , especially at low @xmath4 values , where the @xmath0 background is largest .
a study of the latter could bring some clarification on the prompt - photon puzzle .
section 2 of this paper is devoted to a discussion of the theoretical calculation of the @xmath0 cross sections .
we emphasize the uncertainties associated with the qcd scales , the determination of the fragmentation functions and the importance of the higher - order ( ho ) corrections .
we discuss the comparison of theoretical predictions with data at fixed - target energies in section 3 and with data at higher energies in section 4 .
section 5 contains the conclusions .
at nlo , the inclusive cross section for the hadroproduction of a single hadron @xmath15 , differential in the transverse momentum @xmath4 and the pseudorapidity @xmath16 of @xmath15 , takes the following form : @xmath17 . \end{aligned}\ ] ] the parton densities of the incoming hadrons @xmath18 and @xmath19 are given by @xmath20 and @xmath21 ; the fragmentation of a parton @xmath22 into the hadron @xmath15 is described by the fragmentation function @xmath23 ; @xmath24 is the born cross section of the subprocess @xmath25 ; and @xmath26 is the corresponding ho correction term . in this paper , we shall use the abfow @xcite , cteq @xcite and mrst @xcite parton densities . for the fragmentation functions
, we shall use the parametrization of ref .
these fragmentation functions were derived from fits to charged - pion spectra , but we make the usual assumption , supported by data , that the rate of @xmath0 production is half of that for charged pions .
they are preferable to the parametrizations of ref .
@xcite , which predate the publication of the lep hadronic spectra . expression ( [ 1e ] ) depends on the initial- and final - state factorization scales , @xmath27 and @xmath28 , and on the renormalization scale @xmath29 .
the cross section @xmath30 , calculated to all orders in @xmath31 , is independent of @xmath27 , @xmath28 and @xmath29 .
however , the perturbative series calculated at fixed order in @xmath31 does depend on these scales , the compensation between the variations of @xmath32 , @xmath33 , @xmath34 and @xmath35 and those of @xmath36 being incomplete .
therefore , we shall carefully explore the scale dependence of the cross section ( [ 1e ] ) .
on the one hand , we shall use `` standard scales '' @xmath37 , with @xmath38 varying around @xmath39 , and , on the other hand , we shall fix the scales by using the principle of minimum sensitivity ( pms ) @xcite , @xmath40 which determines the optimal scales @xmath41 , @xmath42 and @xmath43 . an illustration of the scale sensitivity
is presented in fig .
[ fig2.ps ] , where the differential cross section for the inclusive production of single @xmath0 mesons , with @xmath44 gev / c and @xmath45 , through the scattering of 530 gev / c protons on a fixed beryllium target is considered .
specifically , the contours of constant cross section are shown in the @xmath46 plane .
for each set @xmath46 , @xmath29 is determined by the first equality in eq .
( [ 2e ] ) .
as expected , one finds a minimum , corresponding to the optimal choice defined by eq .
( [ 2e ] ) , but one also notices that the area of stability under changes of scales around this optimal point is not very large . it is smaller than in the case of prompt - photon production .
we noticed in several numerical studies that it is not possible to verify the pms criterion ( [ 2e ] ) for too small values of @xmath4 , typically for @xmath47 gev / c .
this indicates that the ho corrections are large and that the qcd predictions are quite unstable under changes of scales .
therefore , we shall emphasize the comparison between theory and experiment at @xmath13 gev / c , although we shall also explore smaller @xmath4 values .
since working in the optimal scheme is rather cumbersome and requires a lot of numerical work , we compare , in general , theory predictions for fixed scales with experiment .
we use the standard choices @xmath48 and @xmath49 . as will be discussed later , the optimal scales , derived from eq .
( [ 2e ] ) , turn out to interpolate between these two choices , depending on the kinematical range considered . besides the scale dependence , there are other sources of uncertainties in eq .
( [ 1e ] ) , which come from the parton densities and the fragmentation functions .
the parton densities are generally determined with good accuracy . however , at small @xmath4 and for scale @xmath50 , the factorization scale may approach the starting scale @xmath51 of the qcd evolution . in this region ,
the parton densities are not constrained by data . to avoid this problem when using the scale @xmath50 , we thus require @xmath4 to be larger than 4.5 gev / c . another source of uncertainty concerns the fragmentation functions .
the quark fragmentation is now well constrained by petra , pep and lep data @xcite . however , the dominant support to eq .
( [ 1e ] ) comes from the large-@xmath52 domain , where the @xmath53 data are scarce .
for instance , numerical studies of eq .
( [ 1e ] ) indicate that the mean @xmath52 value is about 0.85 for @xmath54 gev / c and @xmath55 gev .
there are very few data points in the range @xmath56 ( zero or one point in most data sets ) , and they have large error bars .
therefore , we expect the uncertainty coming from the fragmentation functions to be of the order of a few tens of percents .
it must also be noticed that the gluon fragmentation function is not well determined by @xmath5 data , since it appears there only at nlo . more constraints could , in principle , be obtained from inclusive pion production at large transverse momentum in hadronic collisions .
for example , the inclusive large-@xmath4 charged - hadron cross section has been measured by the ua1 collaboration @xcite in the range @xmath57 gev / c .
these data are compatible with the parametrization of ref .
@xcite , but they are also compatible with the same parametrization after the normalization of the gluon fragmentation function has been increased by 30% to 40% .
( such a large change in the gluon fragmentation function is still compatible with the @xmath5 data . )
this gives us an estimate of the flexibility in the gluon parametrization .
we remark that changing the gluon fragmentation function may affect the slope of the @xmath4 distributions , since the low @xmath58 regions will be more strongly affected . besides these problems related to the @xmath52 behavior of the fragmentation functions
, we should remark that we are led to explore small fragmentation scales ( in particular , when using @xmath59 ) far away from the kinematical regions where the fits are performed . in summary ,
hadronic data require large @xmath52 values and small scales , while the fragmentation functions are ( mostly ) extracted from @xmath5 data at medium @xmath52 values and large scales .
finally , let us mention that we do not use theoretical expression in which the large @xmath60 terms present in the ho corrections are resummed @xcite .
indeed , in order to be coherent , one should also use resummed expressions when extracting fragmentation functions from @xmath5 data .
such an analysis has not been done so far .
it would certainly be very interesting to pursue phenomenological studies in this direction @xcite .
one must , however , keep in mind that the optimization procedure of stevenson and politzer @xcite amounts to a partial resummation of the @xmath60 terms @xcite . for all these reasons , we do not expect a very good agreement between fixed - target data and qcd predictions .
the expected agreement should only be within a few tens of percent . at larger energies , the mean value of @xmath52 is smaller and the sensitivity to the scale variations is reduced .
a better agreement is therefore expected and has been verified in previous studies @xcite .
this point will be discussed later . in this paper ,
all the calculations are performed in the @xmath61 scheme .
the value of @xmath12 in @xmath31 , calculated at two loops , is taken to be equal to the one used in the parton densities .
the data that we are going to discuss are displayed in fig .
[ allpio ] . unlike what was observed in the case of prompt - photon production , the @xmath0 data taken at the same energy are quite compatible with each other , with the exception of isr data at @xmath62 gev , where the r806 @xcite and afs @xcite data disagree for @xmath63 gev / c .
the increase in cross sections as @xmath6 increases at fixed @xmath4 is also visible .
let us emphasize that we only consider @xmath64 , @xmath65 and @xmath66 data in order to reduce the uncertainties due to the parton densities of the incoming hadrons .
we do not study @xmath67 data , but concentrate on proton - induced fixed - target data and isr data in the energy range below @xmath62 gev .
in what follows , we consider only the range @xmath68 gev / c , which is covered by almost all data sets . with this cut
, we also efficiently suppress possible non - perturbative contributions to the cross sections , such as power corrections , intrinsic @xmath14 effects , _
etc_. in fig .
[ fig4ps ] , we show the effects of the variation in the parton densities and scales on the predictions for the wa70 experiment @xcite . for both scale choices , @xmath69 and @xmath50 , we observe the following hierarchy : abfow and cteq5 m lead to the smallest and largest predictions , respectively , while mrst2 leads to an intermediate result .
the result for cteq4 m almost coincides with that for mrst2 and is not shown in fig .
[ fig4ps ] .
this hierarchy may be explained by the different @xmath12 values used in the evaluations of @xmath31 . for four quark flavors ,
these are 230 mev for abfow , 296 mev for cteq4 m , 300 mev for mrst2 and 326 mev for cteq5 m .
when @xmath31 is evaluated with a common value of @xmath12 , the four results become very similar .
the effect of the scale choice is also clearly displayed .
one notices a @xmath4-independent increase by roughly a factor of two when the scale is reduced from @xmath69 to @xmath50 .
this is a typical example for the sensitivity of the theory to the scale choice . as we shall see later ,
the results obtained using optimal scales are close to those obtained with scale @xmath50 .
as expected , the agreement between data and theory is not very good . with small scales ( @xmath70 )
, the theory underestimates the data by some 40% to 50% . in fig .
[ fig5ps ] , ua6 @xmath64 data @xcite are compared with the corresponding qcd predictions .
for this experiment , the qcd predictions for scale @xmath50 undershoot the data by some 30% .
similar remarks hold for the e706 data @xcite . in fig .
[ fig6ps ] , we display data and predictions for @xmath71 gev .
one observes that the curve corresponding to the optimal scales ( labelled pms ) is very close to the one obtained with scale @xmath50 . here
again , the disagreement between theory and data , for scale @xmath50 or optimal scales , is of the order of a few tens of percents .
it is interesting to notice that , with scale @xmath69 , which allows one to extend the predictions down to smaller values of @xmath4 , the theoretical cross section is almost parallel to the data .
therefore , there is no such increase of the ratio of data over theory as was observed in the prompt - photon case ( see fig . [ xtcteqpt2 ] ) .
very similar conclusions can be drawn from the comparison of the e706 data at @xmath72 gev with theory .
finally , in fig . [ fig7ps ] , we show the isr r806 data at @xmath73 gev @xcite and the theoretical predictions obtained with scales @xmath69 and @xmath50 . here again , small scales lead to the best agreement with the data . from this short survey of data and qcd predictions we can draw some preliminary conclusions . as expected , we do not obtain a very good agreement with respect to normalization .
the predictions for scale @xmath50 systematically underestimate the data by some 30% to 50% for all considered experiments . on the other hand ,
the @xmath4 behavior is well reproduced .
one also notices that all data appear to be consistent with each other .
the above discussion is best summarized by displaying on a linear scale the data as differential cross sections in @xmath58 normalized to the theoretical predictions .
the data , normalized to the nlo predictions evaluated using set mrst2 and a common scale set equal to @xmath69 , are displayed in figs .
[ fig8ps ] and [ fig9ps ] .
we notice that the normalized data are mutually compatible within @xmath74 , and that multiplying the theoretical predictions by a common normalization factor of 2.5 will bring data and theory in reasonable agreement .
it is also interesting to remark that the ratio of data over theory is rather flat and shows no such sharp rise at low @xmath58 as was observed in the prompt - photon production experiments ( see fig .
[ xtcteqpt2 ] ) .
these general features are also found when comparing data and theory with the common scale put equal to @xmath50 .
the results are shown in figs .
[ fig10ps ] and [ fig11ps ] for the mrst2 parton distributions .
the major difference between theory and experiment now resides in a normalization factor of about 1.45 . comparing fig .
[ fig11ps ] with fig .
[ fig9ps ] , one notices the different slope of the data - over - theory ratio at small @xmath58 values , corresponding to @xmath75 gev / c , indicating that the theoretical predictions are rather unstable in this region and can not be trusted .
based on the above analysis of fixed - target data ( wa70 , ua6 , e706 , r806 at @xmath76 gev ) , we reach the following conclusions concerning inclusive single-@xmath0 production .
all data appear to be consistent with each other to within @xmath74 .
however , they differ from the fixed - scale theoretical predictions by a rather large normalization factor , namely @xmath77 for scale @xmath69 and @xmath78 for scale @xmath50 . in the latter case , it is worth investigating if this normalization factor could be accounted for , in the theoretical calculations , by a different choice of fragmentation functions , which are still rather flexible in the kinematical range of interest , as explained in section [ theory ] .
this situation is to be compared to that of prompt - photon production , where the scattering of data sets is much larger .
in fact , normalized to the theoretical predictions for fixed scales , the measured e706 rate at @xmath79 gev appears two to four times larger than the corresponding wa70 rate .
another feature distinguishes the @xmath0 spectra from the prompt - photon spectra when they are compared to theory . for scale @xmath69 ,
the data - over - theory ratio for the @xmath0 data is rather flat down to rather low @xmath58 values ( see fig .
[ fig9ps ] ) , while the corresponding ratio for the prompt - photon data is found to exhibit a sharp rise for decreasing @xmath58 . for scale
@xmath50 , a rise is observed for both @xmath0 and prompt - photon data , the rise being sharper in the latter case . this change of behavior signals an instability of the theory : for @xmath0 production at e706 energies , a change of scales introduces a change of slope in the @xmath4 spectrum . for wa70 or ua6 energies ,
the change of scales reduces to a change in the overall normalization .
a final remark concerns the energy variation of the @xmath0 spectrum . by comparing data at different energies in the very same experiment
, we are able to better examine the variation with @xmath6 of the theoretical predictions .
this can be done by looking at figs .
[ fig9ps ] and [ fig11ps ] , where e706 data at @xmath80 gev and 38.8 gev are displayed .
we notice a 30% to 40% difference in normalization , which could indicate that the theoretical predictions calculated with fixed scales are not able to follow the energy dependence of the data .
the use of optimized scales does not improve the agreement with data , since the predictions based on scale @xmath50 or optimized scales give very similar results for all data with energy below @xmath81 gev .
in fig . [ fig13ps ] , we compare two sets of isr data at 63 gev , from the r806 @xcite and afs @xcite collaborations , with theoretical predictions .
in contrast to the results for lower energies , we observe a remarkable stability in the theoretical predictions for @xmath13 gev / c ( @xmath82 ) : all our scale choices lead to similar predictions .
below @xmath83 gev / c , the predictions diverge .
the choice @xmath69 and the use of optimized scales give similar results , while the choice @xmath50 leads to a cross sections which is smaller by a factor of two .
below @xmath84 gev / c , no conclusion is possible because each data set favours a different scale choice . given the statistical relevance of the r806 data , the choice @xmath69 or the use of optimized scales appear to be most appropriate .
the agreement of theory with both data sets is rather good for @xmath85
gev / c , although the theoretical expectations tend to be somewhat higher than the data , a situation which is different from the fixed - target case .
this fact may be related to the observation made above concerning the energy variation of the data - over - theory ratio .
this can be checked , to an extent which is limited by the relatively large error bars , by comparing the fixed - scale theory to the three sets of r806 data at @xmath86 , 44.8 and 62.8 gev .
we find that , at fixed @xmath58 , this ratio tends to decrease when the energy increases , in agreement with the e706 results .
turning to the ua1 data on charged - pion production @xcite , similar comments as above can be made .
the theoretical predictions at large @xmath4 values , above @xmath87 gev / c , are relatively stable and agree well with the data , while , at lower @xmath4 values , the predictions start to diverge , thus bracketing the experimental points .
the fairest conclusion to be drawn from our studies of pion production in hadronic collisions is that the phenomenology of this process is not yet completely understood !
several problems can be identified . on the theoretical side ,
the main difficulty lies in the scale instability , which is significant at low energy but disappears for @xmath88 gev , at least for large - enough @xmath4 values .
hopefully , this instability will be partly removed by resumming the large @xmath89 terms associated with the fragmentation process , very much in the way the resummation of threshold factors improved the predictions of prompt - photon production at large transverse momentum @xcite
. this improvement would be important in order to better understand the energy dependence of the cross sections , which , as we have seen , can not be fully understood when using the fixed - scale approach . in the meantime ,
the choice of optimized scales , which is equivalent to setting the scales equal to @xmath50 at fixed - target energies and to @xmath69 at upper isr energies , seems to be the most appropriate one and is anticipated to give results not too far from the resummed theory ( _ cf . _ the case of prompt - photon production ) . on the phenomenological side
, one needs a second generation of fragmentation functions , based on the resummed approach and also taking account of pion spectra in ( selected ) hadronic collisions . from the @xmath53 data , upon which existing parametrizations are based
, the gluon to pion fragmentation is largely undetermined , and there are very few experimental points to directly constrain the quark to pion fragmentation at very large @xmath52 values .
there is probably enough flexibility in the fragmentation functions to change the size of the predictions for hadronic collisions by 30% or so .
this point is certainly worth a more detailed investigation .
comparing theory with data , the fixed - target data are found to lie systematically above the theory predictions , while the isr data are compatible with or somewhat below the predictions and the ua1 data are in perfect agreement with the predictions .
this problem of energy dependence of the cross section at fixed @xmath58 is particularly manifest in the case of the e706 data .
is this an experimental problem ? is the scale dependence of the theory more involved than expected ? redoing the phenomenology using the resummed approach will probably help to better understand this point .
although confused , the case of @xmath0 production is still much simpler than that of prompt - photon production , where 24 gev and 63 gev data are rather compatible with the predictions , while 30 gev and 40 gev data are much larger . considering only experiments below 40 gev and taking into account the fact that the @xmath0 data taken in those experiments are compatible with each other , this seems to indicate that the systematic errors on prompt - photon production are probably underestimated . in particular ,
the background subtraction necessary to obtain the prompt - photon spectrum should be carefully re - assessed .
it is important for the search of the higgs - boson decay to two photons to understand the production of photons and pions at lhc energies .
given the confused situation at lower energies , after more than twenty years of intense experimental and theoretical efforts , a lot of work remains to be done to achieve this goal .
f. m. borzumati , g. kramer , z. phys .
* c67 * ( 1995 ) 137 ; b. a. kniehl , proceedings of the ringberg workshop on new trends in hera physics , tegernsee , germany , 25 - 30 may 1997 , edited by b. a. kniehl , g. kramer , a. wagner . | we review the phenomenology of @xmath0 production at large transverse momentum in proton - induced collisions .
uncertainties in the next - to - leading - order predictions of quantum chromodynamics are discussed .
the comparison with data reveals that the disagreement between theory and experiment lies essentially in an overall normalization factor .
the situation for @xmath0 production is contrasted with that of prompt - photon production in hadronic collisions .
= 23 truecm = 15 truecm = -0.5 truecm = -2 truecm
@=11 tempcntc citex[#1]#2@fileswauxout tempcnta@tempcntb@neciteaciteforciteb:=#2citeo#1 citeotempcnta > tempcntbciteacitea , tempcnta = tempcntbtempcnta
@=12 to 2 truecm 3 truemm * and next - to - leading - order qcd predictions * 1 truecm p. aurenche,@xmath1 m. fontannaz,@xmath2 j .- ph .
guillet,@xmath1 b. a. kniehl,@xmath3 m. werlen@xmath1 + @xmath1 laboratoire dannecy - le - vieux de physique thorique lapth , + b.p .
110 , f-74941 annecy - le - vieux cedex , france + @xmath2 laboratoire de physique thorique , universit de paris xi , + btiment 210 , f-91405 orsay cedex , france + @xmath3 ii .
institut fr theoretische physik , universitt hamburg , + luruper chaussee 149 , d-22761 hamburg , germany 1 truecm hep - ph/9910252 + lapth-751/99 + lpt - orsay/99/78 + desy 99 - 153 + october 1999 = 20 pt |
stoichiometric euo is a ferromagnetic semiconductor with a curie temperature ( @xmath0 ) of 69 k and a band gap of about 1.2 ev at room temperature.@xcite upon electron doping , the material shows a wealth of spectacular phenomena , including a metal to insulator transition and colossal magnetoresistance , where the change in resistivity can exceed 810 orders of magnitude.@xcite the curie temperature can also be doubled by electron doping@xcite and even almost tripled by pressure.@xcite in the ferromagnetic state the conduction band shows a splitting of about 0.6 ev between the spin up and spin down states leading to an almost 100% spin polarization of the charge carriers in electron doped euo.@xcite these properties make euo a very attractive candidate for fundamental research in the field of spintronics .
euo in thin film form has been studied already in the late 1960s and early 1970s .
use was made of coevaporation of eu and eu@xmath1o@xmath2,@xcite and later , of evaporation of eu in an oxygen atmosphere,@xcite all under _ technical _ vacuum conditions , i.e. , pressures in the range of 10@xmath310@xmath4 mbar .
the preparation of thin films was considered as a convenient alternative synthesis route for euo , alternative to that of the bulk synthesis which required very high temperatures with a delicate phase diagram.@xcite the use of the thin
film preparation route specifically facilitated doping dependence studies using rare earth and transition metal impurities.@xcite after a pause of two decades , a strong renewed interest in euo thin films emerged in recent years .
this time it is the thin film community which is making the attempt to utilize the extraordinary properties of euo for device applications.@xcite part of the motivation also originates from the fact that tremendous progress has been made in preparation technologies , e.g. , molecular beam epitaxy ( mbe ) under ultra high
vacuum conditions , and that new analysis methods have become available , e.g. , synchrotron based spectroscopies .
all in all , these new efforts have culminated in a high point such that euo thin films can be grown epitaxially on si , demonstrating its potential for spintronics applications.@xcite it is remarkable in the recent euo research @xcite that control of the stoichiometry is nevertheless still a serious issue .
many studies reported that eu@xmath5 ions were present in their films and/or that the magnetic moment per f.u .
was not close to the expected 7@xmath6 for a @xmath7 system .
it is not clear in what precision the relative supply rates of oxygen and europium were controlled in these works
. we will show below that this control need not be precise as long as one is in the so called eu distillation condition during growth . also layer by layer growth has to our knowledge never been mentioned , although epitaxy has been often reported .
we therefore set out to do a renewed growth study .
we have chosen for yttria stabilized cubic zirconia ( ysz ) as substrate@xcite : the lattice constant of ysz is 5.142 , practically identical to the 5.144 value for euo at room temperature,@xcite and epitaxy has been reported already.@xcite yet , it was also claimed that control stoichiometry is extremely difficult,@xcite related to the fact that ysz acts as a source of oxygen@xcite during the mbe growth process .
we will show below that we do have achieved full control of the growth process , i.e. , with stoichiometry control and layer by layer epitaxial growth , and that we can use these results as a firm basis for further studies including the doping dependence and the specific use of well defined interfaces .
the euo films were grown in an ultra high vacuum mbe facility with a base pressure of 2@xmath810@xmath9 mbar , maintained by a cryopump .
high purity eu metal from ames laboratory was evaporated from an epi effusion cell with a bn crucible at temperatures between 460@xmath10c and 525@xmath10c .
proper degassing of the eu material ( mostly hydrogen gas ) ensured that during eu evaporation the pressure was kept below 3@xmath810@xmath11 mbar . the eu deposition rate ( 48 / min )
was calibrated using a quartz crystal monitor which was moved to the sample growth position prior and after each growth .
molecular oxygen was supplied through a leak valve , and its pressure ( 416@xmath810@xmath12 mbar ) was monitored using an ion gauge and a mass spectrometer .
the growth was terminated by closing first the oxygen leak valve and then the eu shutter after 30 s. as substrates we used epipolished single crystals of ysz from surfacenet gmbh and cleaved single crystals of mgo from tbl kelpin .
the surface normal of the substrates are all the ( 001 ) .
the lattice constant of ysz is 5.142 , very close to the 5.144 value for euo at room temperature .
the lattice constant of mgo is 4.21 .@xcite prior to growth the substrates were annealed _ in situ _ at @xmath13c in an oxygen atmosphere of @xmath14 mbar for at least 120 min in the case of ysz , and in @xmath15 mbar for at least 60 min for mgo , in order to obtain clean and well ordered substrate surfaces .
the substrates were kept at @xmath16c during growth .
the mbe facility is supplied with the ek35r reflection high energy electron diffraction ( rheed ) system from staib instruments for _ in situ _ and _ online _ monitoring of the growth .
the mbe facility is attached to an ultra high vacuum @xmath17metal photoemission chamber equipped with a scienta ses100 electron energy analyzer and a vacuum generators twin crystal monochromatized al@xmath18 ( @xmath19 = 1486.6 ev ) source for _ in situ _ x ray photoelectron spectroscopic ( xps ) analysis .
the overall energy resolution was set to 0.35 ev and the fermi level @xmath20 was calibrated using a polycrystalline au reference .
the @xmath17metal chamber is also equipped with a vacuum generators scientific t191 rear
view low energy electron diffraction ( leed ) system for further _ in situ _ structural characterization .
the base pressure of the @xmath17metal chamber is 1@xmath810@xmath9 mbar , and all characterizations herein were carried out at room temperature .
the mbe facility is also connected to a separate ultra high vacuum chamber for the evaporation of aluminum as protective capping layer of the air sensitive euo films .
this allows the _ ex situ _ characterizations using x
ray reflectivity ( xrr ) , superconducting quantum interference device ( squid ) , and x ray absorption spectroscopy ( xas ) .
the thickness of the aluminum capping is about 2040 .
the xrr measurements were carried out using a siemens d5000 diffractometer .
the magnetic properties of the films were determined using a quantum design mpms xl7 squid magnetometer .
the xas measurements were performed at the dragon beamline of the national synchrotron radiation research center ( nsrrc ) in taiwan .
the spectra were recorded using the total electron yield method and the photon energy resolution at the eu @xmath21 edges ( @xmath221180 ev ) was set at @xmath23 0.6 ev .
figure 1(a ) shows the rheed photograph of the clean and annealed ysz ( 001 ) before growth , and figs .
1(b)1(e ) show the photographs after 10 min of euo growth .
the eu flux rates were 8.18.2 / min for ( b)(d ) and 4.24.3 / min for ( e)(f ) . the oxygen pressure in the chamber was @xmath24 mbar for ( b ) and @xmath25 mbar for ( c ) and ( e ) .
no oxygen was supplied into the chamber for ( d ) and ( f ) .
the ysz substrate temperature was kept at @xmath16c during growth .
the important result is that the general features of the rheed patterns did not change during growth and that they are very similar to those of the clean ysz for all eu and o growth conditions .
the distance between the streaks of the euo films is identical to that of the pure ysz , confirming that the in
plane lattice constants of euo and ysz are very closely matched .
figure 2 shows the time dependence of the rheed intensity of the specularly reflected beam during the euo growth .
we can clearly observe oscillations which are indicative for a two dimensional ( 2d ) layer by layer or frank van der merwe growth mode .
it is surprising that there are only five to six oscillations for all deposition conditions as indicated in fig . 2 and that these oscillations even exist in the absence of oxygen in the mbe chamber .
it is important to note that the oscillation period does _ not _ depend on the oxygen pressure @xmath26 , thus also in the case of no oxygen in the chamber .
this indicates that the oxygen needed for the formation of euo must also come from the ysz substrate .
the @xmath16c substrate temperature apparently provides sufficient mobility for the oxygen ions to migrate to form at least five or six euo layers .
the oscillation period , which represents a formation of a new atomic single layer , is determined only by the eu flux rate @xmath27 : reducing it by a factor of 2 , from 8.18.2 to 4.24.3 / min , doubles the period , from 25 to 50 s. [ t ] leed photographs for all these films displayed a good single crystallinity .
figure 3 depicts examples for the case of no oxygen in the mbe chamber during growth . also here
we can observe a perfect ( 001 ) surface of the euo rock salt structure , consistent with the rheed results .
the leed photographs were taken at electron beam energies of 213215 ev since lower energies did not provide stable patterns due to charging . to investigate the implications of observing only five to six oscillations
, we also carried out photoemission experiments on those films .
figure 4 , left panel ( a ) , shows the zr @xmath28 and eu @xmath29 core level xps spectra which were collected at normal emission .
it can be clearly seen that the zr signal is reduced when comparing the clean ysz ( bottom curve black ) with the euo covered ysz ( top three curves blue , green , and red ) .
the euo films here were grown without oxygen in the mbe chamber .
different eu flux rates and total time of growth are indicated in the figure .
it is remarkable that the euo covered ysz spectra have very similar zr signals , and also equal eu intensity , despite the fact that the total amount of eu exposure is twice as large in the two middle curves ( green and red ) than in the top one ( blue ) .
this indicates that in the absence of oxygen in the mbe chamber , the growth of euo is limited to five to six monolayers only and that the rest of the deposited eu metal is re evaporated back into the vacuum ( the substrate temperature is @xmath30c ) . in other words ,
the sticking coefficient for eu after the completion of five to six monolayers is reduced to zero , suggesting that oxygen transport through euo is much more difficult than in ysz .
it is also important to investigate the chemical state of the eu .
figure 4 , right panel ( b ) , depicts the o @xmath31 and eu @xmath32 valence band spectra together with the zr @xmath33 , o @xmath34 , and eu @xmath35 core levels .
the eu @xmath32 lineshape in all the films is very characteristic for a eu@xmath36 system .
the multiplet structure typical for eu@xmath5 is not visible .
one can also observe that the o @xmath31 spectrum at 610 ev binding energy for ysz is converted into the o @xmath31 valence band at 47 ev typical for euo .
@xcite all these demonstrate that only euo has been formed , free from eu@xmath1o@xmath2 or eu@xmath2o@xmath37 contaminants .
this also means that ysz can only oxidize eu into the 2 + state , and definitely not into the 3 + .
we have also carried out take off angle dependent xps experiments on the films .
figure 5 shows the zr @xmath28 and eu @xmath29 core level xps spectra of one of the euo films of fig .
4 ( a ) collected at grazing emission , i.e. , @xmath38 with respect to the surface normal , and at normal emission .
one can clearly see that the eu signal is not significantly reduced but the zr signal has almost disappeared in the grazing emission geometry .
since grazing emission means more surface sensitivity , this result confirms not only that the euo film is on top of the ysz substrate with negligible intermixing of the cations but also that the film is closed and flat .
figure 6 depicts the zr @xmath28 and eu @xmath29 core level spectra ( left panel ) , and the zr @xmath33 , o @xmath34 , and eu @xmath35 core level together with o @xmath31 and eu @xmath32 valence band spectra ( right panel ) of euo films grown with the supply of oxygen in the mbe chamber .
various oxygen pressures have been used as indicated in the figure .
the eu flux rate and the deposition time are identical for these films .
one can clearly observe that the zr @xmath28 signal is getting smaller when the oxygen pressure is increased , indicating that the thickness of the euo film becomes larger .
the lineshapes of the eu @xmath29 and eu @xmath32 levels are those of divalent eu for pressures up to 12@xmath810@xmath12 mbar . for a pressure of 12@xmath810@xmath12 mbar or higher ,
however , the eu @xmath29 and eu @xmath32 spectral shapes start to change and show characteristics which indicate the presence of trivalent eu . apparently , for a eu flux rate of 8.2
/ min and substrate temperature of @xmath30c , 1012@xmath810@xmath12 mbar is the critical oxygen pressure below which euo films can be made on ysz ( 001 ) free from any eu@xmath1o@xmath2 or eu@xmath2o@xmath37 type of impurity phases .
we have also measured the photoemission spectra of these euo films under grazing take off angle conditions .
again , the eu @xmath29 core level and eu @xmath32 valence band spectra are all 2 + as long as the oxygen pressures in the mbe chamber are below the critical 1012@xmath810@xmath12 mbar value .
one example is shown in fig . 7 , where an oxygen pressure of @xmath24 mbar was used .
this figure demonstrates that also the surface region of the films is free from eu@xmath5 species .
it is interesting to compare the grazing with the normal emission spectra and also with the spectra displayed in fig .
one can clearly see in fig .
7 that the zr @xmath28 signal is very much suppressed in grazing emission , even more suppressed than in the grazing emission spectrum of fig .
this shows that the euo films grown with the supply of oxygen in the mbe chamber are thicker , a not so surprising and yet very consistent observation since without oxygen we have found that the euo film growth is limited to five to six monolayers .
the following picture can now be drawn about the initial stages of growth of euo film on ysz ( 001 ) . in case
there is no oxygen in the mbe chamber , ysz is supplying all the oxygen required to form euo .
the film is perfectly free from eu@xmath5 species . in case oxygen
is present in the mbe chamber , ysz is supplying only the amount of oxygen that is needed to complete the formation of euo .
this also explains why the growth rate of the first five to six monolayers is determined only by the eu flux rate and is totally independent of the supply of oxygen pressure in the mbe chamber , see fig .
it is important that the pressure is kept below the critical value of 1012@xmath810@xmath12 mbar as we will discuss in more detail in sec .
[ four ] . based on the comprehensive set of rheed , leed , and xps data ,
including the rheed intensity oscillations , we have now demonstrated that euo thin films can be grown epitaxially in a layer by layer fashion with good control of its chemical state .
the supply of oxygen from the ysz does not do any harm , and in fact , it can be utilized as a welcoming method to calibrate the eu flux rate accurately , e.g. , the 8.18.2 / min from the quartz crystal monitor corresponds to the growth of one monolayer euo per 25 s.
having shown that the initial stages of growth of euo on ysz ( 001 ) can be made quite perfect , we now investigate whether thicker euo films can be prepared while keeping the epitaxy and , especially , the layer by layer growth mode .
we therefore have grown films for longer deposition times , e.g. , between 100 and 200 min , using a series of finely intervalled pressures for the oxygen , e.g. , 4 , 8 , 10 , 12 , and 16 @xmath39 mbar .
the eu flux rates were kept at 8.08.3 / min .
the resulting thickness of the films varies between roughly 300 and 800 as will be discussed later .
the rheed and leed results are plotted in fig .
one can clearly see that excellent epitaxial growth has been achieved for ( a ) 4 , ( b ) 8 , and ( c ) 10 @xmath39 mbar oxygen pressures . for 12 @xmath39 mbar ( d ) or higher pressures , however , the appearance of additional spots in the rheed indicates that the surface structure starts to change , and the absence of a pattern in the leed even suggests appreciable surface roughness .
the leed photographs were taken at electron beam energies of 368370 ev since lower energies did not provide stable patterns due to charging . as discussed in sec [ three ] ,
the initial five to six oscillations of the specular reflected rheed beam intensity are a unique feature for the initial stages of euo growth on the ysz ( 001 ) .
these initial oscillations do always occur , i.e. , independent of the oxygen pressure in the mbe chamber , unless the pressure exceeds a critical value above which trivalent eu species are formed .
remarkable is that no more oscillations can be observed beyond these five to six when growing thicker films .
this is the case for a wide range of oxygen pressures .
there is one exception : for a pressure of @xmath40 mbar , we were able to see further rheed intensity oscillations .
figure 9 shows that the initial five to six oscillations are then followed by at least 50 more oscillations .
it demonstrates that a layer by layer growth mode for euo is possible during the sustained growth .
interestingly , the oscillation period during the sustained growth is similar and yet a little bit larger than during the initial stages of growth : 30 s against 25 s. apparently , the oxygen pressure must be close to and yet a little less than the critical value in order to maintain the layer by layer growth mode , while the eu flux determines the growth rate in the initial stages , i.e. 8.08.2 / min eu flux corresponding to 25 s. per euo layer , it is the limited oxygen supply from the mbe environment which dictates the speed during the sustained growth , i.e. , to 30 s per euo layer . to elucidate further the growth process , we have measured the thickness of the films using _
ex situ _ xrr measurements .
since euo deteriorates rapidly under ambient conditions , the films need to be capped . to this end ,
an aluminum layer with a thickness of 2040 has been evaporated on top of the euo .
this thickness turns out to be sufficient for the aluminum to be a good protective overlayer as will be discussed later .
figure 10 exhibits the xrr profile of the euo film which was grown at an oxygen pressure of 4@xmath810@xmath12 mbar for 200 min .
the corresponding rheed and leed patterns of the euo film are displayed in fig .
8(a ) . from the period of interference fringes
, we deduce that the thickness of the euo film is about 350 .
thicknesses of the other films are also determined from their xrr profiles .
the results are displayed in fig .
11 , where we plot the thickness against the product of oxygen pressure and total deposition time . for oxygen pressure
up to @xmath41 mbar we can observe a clear and direct linear relationship between them , strongly suggesting that the thickness is determined by the amount of oxygen incorporated .
in other words , the growth is limited by the amount of oxygen made available .
this in turn means that the eu flux rate is higher than necessary and that the excess eu must be re evaporated into the vacuum .
figure 11 essentially confirms the distillation process needed to maintain good control of the stoichiometry as reported in our earlier studies by steeneken _
et al_.@xcite and tjeng _ et al_.@xcite we now investigate to what extent the growth conditions affect the magnetic properties of the euo films using a squid magnetometer .
the results are shown in fig .
the films grown with 4 , 8 , 8.5 , and 10 @xmath39 mbar oxygen pressures all have a curie temperature of 69 k with a magnetic moment of 7 @xmath6/f.u . as expected for a @xmath7 system .
the inset shows the field dependence of the magnetization at 5 k for the film grown with a @xmath42 mbar oxygen pressure . here
one can observe a hysteresis behavior with a saturation magnetization of 7@xmath6 .
these results are in agreement with the rheed and leed results as displayed in figs .
8(a)8(c ) , in the sense that the proper ferromagnetic properties are always maintained as long as good epitaxial growth is also achieved .
on the other hand , films grown with 12 and @xmath43 mbar have completely lost their ferromagnetic properties .
it is remarkable that exceeding the @xmath40 mbar value just a little bit causes such a dramatic change .
this very abrupt change is also consistent with the rheed and leed results as displayed in fig .
8 ( d ) .
considerable film roughness starts to develop for oxygen pressures higher than @xmath40 mbar .
we would like to infer that having only a @xmath0 of about 69 k is not a sufficient characteristic to conclude that the film is homogeneous and stoichiometric .
one also needs to establish that the film has a full saturation magnetization of 7@xmath6 .
the measurement of the magnetic properties can therefore serve as a critical test for the growth conditions and in particular , the oxygen stoichiometry of the euo films .
we have also performed _
ex situ _ soft xas measurements at the eu @xmath21 edges to examine the integrity of the euo films after capping with the aluminum overlayer .
figure 13 depicts the xas spectra together with the theoretical spectra for eu@xmath36 ( top ) and eu@xmath5 ( bottom).@xcite it is clear that the eu spectra are very similar to the theoretical spectrum for eu@xmath36 , meaning that the euo films with 4 , 8 , and 10 @xmath39 mbar oxygen pressures are completely free from eu@xmath5 species .
this in turn implies that an aluminum overlayer as thin as 2040 works well to protect the euo films against air , contrary to the claims made elsewhere that one needs very thick capping layers.@xcite we attribute this to the fact that the epitaxial growth of euo on ysz ( 001 ) yields such a smooth film so that a very thin aluminum film is sufficient to make a closed capping overlayer .
the smoothness of the films as well as the complete absence of eu@xmath5 impurities forms a good starting point for the fabrication of well defined interfaces with other metals or oxide materials , thereby opening up new opportunities to study or even generate new phenomena related to interface physics .
we conclude that the critical oxygen pressure is around 1012@xmath39 mbar for a 8.08.3 / min eu flux rate . only below this pressure
one has the distillation process taking place so that good epitaxial growth can be achieved with the proper stoichiometry and ferromagnetic properties .
apparently , layer by layer growth can be obtained only if one is close to , but not exceeding , the critical pressure .
[ b ] having understood the growth process of euo on ysz ( 001 ) , and having found the recipe to obtain excellent epitaxial growth of euo on ysz ( 001 ) , we now turn our attention to the preparation of euo on mgo ( 001 ) .
mgo as a substrate for euo is interesting since several studies@xcite have reported good epitaxial growth despite the very large lattice mismatch of about 20% .
the rheed and leed photographs of euo films grown on mgo using different oxygen pressures are displayed in fig .
euo films with oxygen pressure below 10@xmath810@xmath12 mbar all show excellent epitaxial growth .
the relationship between the distances of the mgo and euo streaks is consistent with the ratio of their lattice constants of 1.22 . on the other hand ,
a higher oxygen pressure creates additional spots in the rheed image and even causes the leed pattern to disappear .
similar to the euo on ysz case , this indicates that 1012@xmath810@xmath12 is the critical value for the oxygen pressure above which one no longer gets crystalline and stoichiometric euo .
our attempts to observe rheed intensity oscillations of euo films grown on mgo ( 001 ) were yet unsuccessful .
the rheed streaks and specular spot suddenly disappeared right after the deposition of euo has been initiated . in approximately 30 s , new streak lines appear , whose spacing conforms to the euo lattice parameter
. however , the specular spot was never recovered .
a more detailed growth study is required to determine the growth process of euo on mgo ( 001 ) . to investigate the quality of these euo films in terms of their magnetic properties ,
we have performed squid measurement for films grown with oxygen pressures of 8 , 10 , and 12@xmath810@xmath12 mbar .
the results are shown in fig . 15 .
similar to the euo on ysz case , the films that were grown below 10@xmath810@xmath12 mbar have a curie temperature of 69 k with a magnetic moment close to 7@xmath6/f.u .
conversely , the film grown under a higher oxygen pressure has completely lost its ferromagnetic character .
the small peak at roughly 5 k indicates the typical antiferromagnetic ordering temperature of eu@xmath2o@xmath37 , meaning that the dramatic loss of the ferromagnetism for the growth using slightly above the critical oxygen pressure is due to the presence of eu@xmath5 species .
we have successfully grown epitaxial and highly stoichiometric euo films on ysz ( 001 ) .
the initial stages of growth involve the limited supply of oxygen from the ysz substrate , but the euo stoichiometry can still be well maintained .
we have also observed rheed intensity oscillations during the sustained stages of growth , which demonstrate that the layer by layer growth mode can be achieved for euo films on ysz ( 001 ) .
the euo films were sufficiently smooth so that capping with a thin layer of aluminum enabled _ ex situ _ experiments free from trivalent eu species . the excellent epitaxial growth of euo films is always accompanied by
equally excellent ferromagnetic properties : a saturation magnetization of 7@xmath6 and a @xmath0 of 69 k. we have also confirmed that the use of the eu distillation process during the mbe assisted growth enables the consistent achievement of stoichiometry .
the layer by layer growth , the smoothness of the film , and the excellent stoichiometry provide an excellent basis for the fabrication of well defined interfaces for further studies .
we would like to thank lucie hamdan for her skillful technical and organizational assistance in preparing the experiment .
we acknowledge the nsrrc staff for providing us with beam time .
the research in cologne is supported by the deutsche forschungsgemeinschaft through sfb 608 .
g. m. roesler , jr . , m. e. filipkowski , p. r. broussard , y. u. idzerda , m. s. osofsky , and r. j. soulen , jr . , in _ superconducting superlattices and multilayers _ , edited by i. bozovic , ( spie bellingham , 1994 ) , vol . 2517 , pp .
285 - 290 .
a. schmehl , v. vaithyanathan , a. herrnberger , s. thiel , c. richter , m. liberati , t. heeg , m. rckerath , l. f. kourkoutis , s. mhlbauer , p. bni , d. a. muller , y. barash , j. schubert , y. idzerda , j. mannhart , and d. g. schlom , nat . mater .
* 6 * , 882 ( 2007 ) . | we have succeeded in growing epitaxial and highly stoichiometric films of euo on yttria stabilized cubic zirconia ( ysz ) ( 001 ) .
the use of the eu distillation process during the molecular beam epitaxy assisted growth enables the consistent achievement of stoichiometry .
we have also succeeded in growing the films in a layer by layer fashion by fine tuning the eu vs. oxygen deposition rates .
the initial stages of growth involve the limited supply of oxygen from the ysz substrate , but the euo stoichiometry can still be well maintained .
the films grown were sufficiently smooth so that the capping with a thin layer of aluminum was leak tight and enabled _ ex situ _ experiments free from trivalent eu species .
the findings were used to obtain recipes for better epitaxial growth of euo on mgo ( 001 ) . |
macroscopic quantum coherence manifests itself in many specific phenomena .
one of them is a non - dissipative drag that takes place in superfluids and superconductors .
the non - dissipative drag , also known as the andreev - bashkin effect , was considered , for the first time , in ref .
@xcite , where a three velocity hydrodynamic model for @xmath0he-@xmath1he superfluid mixtures was developed .
it was shown that superfluid behavior of such systems can be described under accounting the `` drag '' term in the free energy .
this term is proportional to the scalar product of the superfluid velocities of two superfluid components .
a similar situation may take place in mixtures of superfluids of @xmath2 and @xmath3 pairs in liquid @xmath0he in the @xmath4-phase @xcite . among other objects , where the non - dissipative drag may be important ,
are neutron stars , where the mixture of neutron and proton cooper pair bose condensates is believed to realize @xcite .
the possibility of realization of the non - dissipative drag in superconductors was considered in @xcite .
the non - dissipative drag in bilayer bose systems was treated microscopically in @xcite for a special case of two equivalent layers of charged bosons .
the case of a bilayer system of neutral bosons was studied in @xcite in the limit of small interlayer interaction .
the most promising systems where the non - dissipative drag can be observed experimentally are two - component alkali metal vapors . in such systems
the interaction between atoms of different species is of the same order as the interaction between atoms of the same specie and the effect is expected to be larger than in bilayers . in bose mixtures the components are characterized by different densities , different masses of atoms and different interaction parameters . in this paper
we consider such a general case and obtain an analytical expression for the drag current for zero and finite temperatures .
in the system under consideration the drag force influences the dynamics of atoms in the drag component in the same manner as the vector potential of electromagnetic field influences the dynamics of electrons in superconductors .
in particular , in neutral superfluids with josephson links the drag effect may induce the gradient of the phase of the order parameter in the bulk and , as a consequence , control the phase difference between weakly coupled parts of the system .
therefore , on can expect that the effect reveals itself in a modification of josephson oscillations between weakly coupled bose gases . in this paper
we discuss possible ways for the observation such a modification .
we consider the bose gas confined in a toroidal trap with two josephson links . in the fock regime
@xcite the low energy dynamics of the system can be described by the qubit model of general form ( the model , where all three components of the pseudomagnetic field can be controlled independently ) .
the parameters of the qubit hamiltonian depend on the drag factor .
the measurement of the state of the qubit under controlled evolution allows to observe the effect caused by the non - dissipative drag and determine the drag factor . in this paper
we consider two particular schemes of the measurement . in the first scheme
one should determine the time required to transform a reproducible initial state to a given final state .
in the second scheme the geometrical ( berry ) phase should be detected . in sec .
[ sec2 ] the microscopic theory of the non - dissipative drag in two - component bose gases is developed . in sec .
[ sec3 ] a model of the bose - einstein qubit subjected by the drag force is formulated and the schemes of measurement of the drag factor are proposed .
conclusions are given in sec .
let us consider a uniform two - component atomic bose gas in a bose - einstein condensed state .
we will study the most general situation where the densities of atoms in each component are different from one another ( @xmath5 ) , the atoms of each components have different masses ( @xmath6 ) and the interaction between atoms is described by three different scattering lengths ( @xmath7 ) .
the hamiltonian of the system can be presented in the form @xmath8 where @xmath9\nabla\hat{\psi}_i({\bf r})\ ] ] is the kinetic energy , @xmath10 is the energy of interaction , @xmath11 and @xmath12 are the interaction parameters , and @xmath13 are the chemical potentials . for the further analysis it is convenient to use the density and phase operator approach ( see , for instance , @xcite ) .
the approach is based on the following representation for the bose field operators @xmath14 \sqrt{n_i+\hat{n}_i({\bf r})},\ ] ] @xmath15,\ ] ] where @xmath16 and @xmath17 are the density and phase fluctuation operators , @xmath18 are the @xmath19-number terms of the phase operators , which are connected with the superfluid velocities by the relation @xmath20 . in what follows
we specify the case of the superfluid velocities independent of @xmath21 . substituting eqs .
( [ 3 ] ) , ( [ 4 ] ) into eq .
( [ 1 ] ) and expanding it in series in powers of @xmath16 and @xmath22 we present the hamiltonian of the system in the following form @xmath23 in ( [ 5 ] ) the term @xmath24 + \gamma_{12}n_1 n_2\bigg)\ ] ] does not contain the operator part . here
@xmath25 is the volume of the system .
the minimization conditions for the hamiltonian @xmath26 yield the equations @xmath27 under the conditions ( [ 5d ] ) the terms , linear in the density fluctuation operators , vanish in the hamiltonian . taking into account the @xmath28 , we find that the terms , linear in the phase fluctuation operators , vanish in the hamiltonian as well .
the part of the hamiltonian quadratic in @xmath22 and @xmath16 operators reads as @xmath29 \cr + \frac{\hbar { \bf v}_i}{2}\big(\hat{n}_i({\bf r})\nabla\hat{\varphi}_i({\bf r } ) + [ \nabla\hat{\varphi}_i({\bf r})]\hat{n}_i({\bf r})\big)\cr + \frac{i\hbar^2}{2m_i}\big([\nabla\hat{n}_i({\bf r})]\nabla\hat{\varphi}_i({\bf r})- [ \nabla\hat{\varphi}_i({\bf r})]\nabla\hat{n}_i({\bf r})\big ) + \frac{\gamma_{ii}}{2}\big(\hat{n}_i({\bf r})\big)^2\bigg\ } + \gamma_{12}\hat{n}_1({\bf r})\hat{n}_2({\bf r } ) \bigg).\end{aligned}\ ] ] the quadratic part of the hamiltonian determines the spectra of the elementary excitations .
hereafter we will neglect the higher order terms in the hamiltonian ( [ 5 ] ) .
these terms describe the scattering of the quasiparticles and they can be omitted if the temperature is much smaller than the temperature of bose - einstein condensation .
let us rewrite the quadratic part of the hamiltonian in terms of the operators of creation and annihilation of the elementary excitations .
as the first step , we use the substitution @xmath30,\ ] ] @xmath31,\ ] ] where operators @xmath32 , @xmath33 satisfy the bose commutation relations .
here @xmath34 is the spectrum of free atoms , and @xmath35 is the spectrum of the elementary excitations at @xmath36 and @xmath37 . the substitution ( [ 6 ] ) , ( [ 7 ] ) reduces the hamiltonian ( [ 5c ] ) to the form quadratic in @xmath32 and @xmath33 operators : @xmath38 \cr + \sum_{\bf k } g_k\left[b^+_{1}({\bf k } ) b_{2}({\bf k})+b_{1}({\bf k } ) b_{2}({\bf -k})+h.c.\right].\end{aligned}\ ] ] here @xmath39 and @xmath40 the hamiltonian ( [ 8 ] ) contains non - diagonal in bose creation and annihilation operator terms and it can be diagonalized using the standard procedure of u - v transformation @xcite .
the result is @xmath41,\ ] ] where @xmath42 and @xmath43 are the operators of creation and annihilation of elementary excitations . the energies @xmath44 satisfy the equation @xmath45 where @xmath46 @xmath47 and @xmath48 is the identity matrix .
the densities of superfluid currents in two components can be obtained from the relation @xmath49 where @xmath50 is the free energy of the system . here
the quantity @xmath51 is defined as the density of the mass current . the free energy of the system , described by the hamiltonian ( [ 5 ] ) , is given by the formula @xmath52
+ t \sum_{{\bf
k}}\sum_{\lambda=\alpha,\beta } \ln\left[1-\exp\left(-\frac{{\cal e}_{\lambda } ( { \bf k})}{t}\right ) \right].\ ] ] the second term in ( [ 201n ] ) is the energy of the zero - point fluctuations and the third term is the standard temperature dependent part of the free energy for the gas of noninteracting elementary excitations .
we specify the case of small superfluid velocities ( much smaller than the critical ones ) . in this case
the currents can be approximated by the expressions linear in @xmath53 .
to obtain these expressions we will find the free energy as series in @xmath54 , neglecting the terms higher than quadratic . at @xmath55 the equation ( [ n1 ] )
is easily solved and the spectra are found to be @xmath56 as required in the procedure @xcite , we take positive valued solutions of eq .
( [ n1 ] ) .
the energies ( [ n3 ] ) should be real valued quantities .
this requirement yields the common condition for the stability of the two - component system : @xmath57 .
if this condition were not fulfilled , spatial separation of two components ( at positive @xmath58 or a collapse ( at negative @xmath59 ) would take place . at nonzero superfluid velocities we present the solutions of eq .
( [ n1 ] ) as series in @xmath53 : @xmath60 @xmath61 note that at @xmath62 the spectra ( [ n4 ] ) , ( [ n5 ] ) are reduced to common expressions for the energies of quasiparticles in a moving condensate : @xmath63 .
using eqs .
( [ 201n]),([n4 ] ) and ( [ n5 ] ) we obtain the following expression for the free energy @xmath64,\ ] ] where @xmath65 does not depend on @xmath53 . in ( [ 28a ] ) @xmath66 are the mass densities , the quantities @xmath67,\ ] ] @xmath68\ ] ] describe the thermal reduction of the superfluid densities , and the quantity @xmath69,\end{aligned}\ ] ] which we call the `` drag density , '' yields the value of redistribution of the superfluid densities between the components . in eqs .
( [ n6])-([n7 ] ) @xmath70^{-1}$ ] is the bose distribution function . using eqs .
( [ 101n ] ) , ( [ 28a ] ) we arrive to the following expressions for the supercurrents @xmath71 @xmath72 one can see that at nonzero @xmath73 the current of one component contains the term proportional to the superfluid velocity of the other component .
it means that there is a transfer of motion between the components .
in particular , at @xmath74 the current in the component 1 ( @xmath75 ) is purely the drag current . since @xmath73 is the function of @xmath76 ( see eqs .
( [ n7 ] ) and ( [ n3 ] ) ) the drag current does not depend on the sign of the interaction between the components .
( [ n7 ] ) is the main result of the paper .
this equation yields the value of the drag for the general case of two - component bose system with components of different densities , different masses of atoms , different interaction parameters , and for zero as well as for nonzero temperatures .
moreover , this equation is valid not only for the point interaction between the atoms , but for any central force interaction . in the latter case
the interaction parameters @xmath77 in eq .
( [ n7 ] ) and in the spectra ( [ n3 ] ) , ( [ 7a ] ) should be replaced with the fourier components of the corresponding interaction potentials . to estimate the absolute value of the drag we , for simplicity , specify the case @xmath78 ,
that is realized when two components are two hyperfine states of the same atoms . at @xmath79 eq .
( [ n7 ] ) is reduced to @xmath80 where @xmath81 is the density of states for free atoms , and @xmath82 the integral in ( [ n11 ] ) can be evaluated analytically . to present the answer in a compact form it is convenient to introduce the dimensionless parameters @xmath83 ( @xmath84 and @xmath85 ) using these notations we have @xmath86 { n_1 a_{11}^3 n_2 a_{22}^3}\frac { \eta}{\sqrt{\kappa}}f(\kappa,\eta),\ ] ] where @xmath87 direct evaluation of eq .
( [ n11c ] ) shows that at allowed @xmath88 and @xmath89 ( @xmath84 and @xmath85 ) the factor @xmath90 is almost the constant ( the range of variation of @xmath50 is @xmath91 $ ] ) and one can neglect the dependence of @xmath50 on the parameter of the system . at @xmath92
we obtain from ( [ n11b ] ) the following approximate relation @xmath93 if the density of one component is much larger than of the other and @xmath94 , the `` drag density '' is approximated as @xmath95 one can see that the `` drag density '' is proportional to the square root of the gas parameter .
it means that the drag effect is larger in `` less ideal '' bose gases .
the temperature dependence of the `` drag density '' at small @xmath96 can be evaluated analytically from eq .
( [ n7 ] ) using the linear approximation for the spectra of the excitations .
it yields @xmath97 , where @xmath98 and the factor @xmath99 is positive .
numerical evaluation of the sum over @xmath100 in eq .
( [ n7 ] ) shows that the analytical approximation is valid only at @xmath101 . at @xmath102
the `` drag density '' decreases much slower under increase of the temperature . as an example , the dependence of @xmath103 at @xmath104 , @xmath105 and @xmath106 is shown in fig .
[ fig1 ] .
dependence of the `` drag density '' on the temperature.,height=226 ] now let us discuss how the drag effect can reveal itself in a real physical situation .
if one deals with the stationary superflow one implies that it is the circulating superflow , e.g. , the tangential superflow in a hole cylinder . in such a case
the superfluid velocities satisfy the onsager - feymnan quantization condition @xmath107 where the vorticity parameters @xmath108 are integer .
then , the drag effect can be understood as the appearance of the circulating current in the drive component ( e.g. specie 1 ) , when the circulation of the superfluid velocity of the drive component ( e.g. specie 2 ) is fixed ( @xmath109 ) .
the current of the specie 1 ( [ 22 ] ) depends on the superfluid velocities of the both species and if the superfluid velocity of the drag component directed anti - parallel to the superfluid velocity of the drive component the current of the drag component might vanish .
but since the velocities are quantized it may happen only under certain special conditions ( see below ) .
the superfluid velocity of the drag component is determined by that at fixed @xmath110 the free energy ( [ 28a ] ) has a minimum with respect to discrete values of @xmath111 ( where the @xmath112 is the radius of the contour in ( [ d1 ] ) ) .
depending on the value of the parameter @xmath113 several possibilities can be realized . at @xmath114
the minimum of the energy ( [ 28a ] ) corresponds to @xmath115 ( and @xmath116 ) . in this case
the current of the drag component is directed along the drive current and it is proportional to the drag density . at @xmath117 (
@xmath118 is natural ) the value @xmath119 minimizes the energy . in this case two terms in eq .
( [ 22 ] ) compensate each other and the current in the drag component vanishes . at half - integer
@xmath120 the degenerate situation takes place : two state ( with co - directed currents , and counter - directed currents ) have the same energy . at @xmath121 the state with counter - directed currents
gains the energy and at @xmath122 the co - directed currents are energetically preferable . in the latter two cases the nonzero vorticity of the drag component ( @xmath123 ) is also induced .
this behavior is analogous the behavior of a superconducting ring in a magnetic field .
we note that since @xmath124 , the most realistic case is @xmath114 when the simple picture of the transfer of part of the motion from the drive to the drag component takes place . in this study
we have concentrated on the analytical derivation of the drag effect in the uniform bose gases .
the consideration of the non - uniform case requires the solution of the eigenvalue problem for the elementary excitations in the two - component bose gas in the external potential .
but even for the simplest case of a spherically symmetric trap this problem can be solved analytically only in the long - wavelength limit and the thomas - fermi approximation @xcite ( the spectrum of elementary excitations in one - component bose gases was obtained analytically for a number of potentials but also in the same limit @xcite ) . since the main contribution to the drag density comes from the excitations with the wave vectors of order of the healing length ( see ( [ n7 ] ) ) , the rigorous analysis of the drag effect can be done only numerically .
nevertheless , in the tomas - fermi situation the drag effect can be evaluated basing on the following arguments .
when the linear size of the bose cloud is much larger than the healing length , the spectrum of the excitations at the wave vectors of order or higher than the inverse healing length is well described by the quasi uniform approximation .
therefore , the drag effect can be described by the same equations , as in the uniform case with the only modification that the quantities @xmath125 and @xmath126 , and , correspondingly , @xmath127 , @xmath128 , @xmath73 and @xmath129 in eqs .
( [ n6])-([23 ] ) are understood as functions of coordinates . at an arbitrary symmetry of the trap potential the superfluid velocity of the drag component
can not be equal to zero in each point .
indeed , in general case of space dependent @xmath127 , @xmath128 , and @xmath73 the velocity field @xmath130 can not satisfy two independent continuity conditions @xmath131=0 $ ] and @xmath132 . to analyze this case
one should find the velocity fields @xmath133 and @xmath130 that satisfy the continuity conditions and the quantization conditions . to illustrate this point
let us consider a simple example of a trap having the shape of a hollow cylinder with the densities that depend only on the polar angle @xmath134 .
we will seek the velocity fields that do not have radial components .
then , the eqs .
( [ 22 ] ) , ( [ 23 ] ) , written for the tangential components of the currents and the velocities , can be presented in the matrix form @xmath135 where @xmath136 with @xmath137 . due to the continuity conditions the current @xmath138 and @xmath139 in ( [ mm1 ] )
do not depend on @xmath134 .
according to eq .
( [ mm1 ] ) the velocities @xmath140 and @xmath141 are connected with the currents by the equation @xmath142 integrating eq .
( [ mm3 ] ) over @xmath134 and taking into account the quantization conditions ( [ d1 ] ) we obtain the equation for the currents @xmath143 where @xmath144 if a given vorticity of the drive component @xmath110 is not very large the minimum of energy is reached at @xmath115 . in the latter case the solution of eq .
( [ mm4 ] ) in the leading order in @xmath73 yields the following expression for the current of the drag component @xmath145 one can see that if at some @xmath134 the density @xmath146 has a sharp minimum the first factor in denominator in eq .
( [ ddd ] ) becomes large . on the other hand ,
the integral in numerator is not very sensitive to lowering of @xmath146 ( see eqs . ( [ 1002 ] ) ) .
thus , in a system with a `` bottle neck '' in the drag component the drag current decreases strongly and the main consequence of the drag effect is the emergence of the gradient of the phase of the order parameter of the drag component .
similar situation takes place in a system with a weak link .
the latter case is analyzed in the next section . in the uniform case
( [ ddd ] ) is reduced to @xmath147 . to complete the discussion we emphasize that the crossed term ( @xmath148 ) in the free energy ( [ 28a ] ) ( and , consequently , the drag terms in the currents ( [ 22 ] ) , ( [ 23 ] ) ) comes only from the second and third terms in eq .
( [ 201n ] ) .
consequently , the drag effect considered in this paper is solely by the excitations . at the mean field level of approximation ( which can be also formulated in terms of the gross - pitaevsky equation )
the effect does not appear , while the coupling between the components is also present at that level of approximation .
we would note that at the mean field level the drag effect of another type may emerge .
that effect takes place in the case when one of the species is subjected by an asymmetric rotating external potential ( see , for instance , @xcite , where such an effect has been studied with reference to the system of two coupled traps ) .
it is known that bose systems in the bose - einstein condensed state may demonstrate josephson phenomenon @xcite .
it this paper we consider the external josephson effect that takes place in two - well bose systems .
it was shown in @xcite that in such systems one can realize the situation , when two states , that differ in the expectation value of the relative number operator , can be used as qubit states . to include the drag force into the play we consider the following geometry .
let our two - component system is confined in a toroidal trap and the bose cloud of the component 1 ( the drag component ) is situated inside and overlaps with the bose clouds of the component 2 ( the drive component ) .
such a situation can be realized if @xmath149 .
deforming the confining potential one can cut the drag component into two clouds of a half - torus shape ( separated by two josephson links ) leaving the bose cloud of the drive component uncutted ( fig .
[ fig2 ] ) . in what follows we use the following notations : @xmath150 is the large radius of the toroidal trap , @xmath151 and @xmath152 are the small radiuses of the toroidal bose clouds of the drag and the drive components , correspondingly . schematic shapes of bose clouds for the drag ( top figure ) and drive ( bottom figure ) components .
the drag component is situated inside and overlaps with the drive component.,height=226 ] rotating this trap one can excite a superflow in the drive component .
after the rotation be switched off there will be a circulating superflow in the drive component and no superflow in the drag component ( at negligible small josephson coupling ) .
the superfluid velocity of the drive component is @xmath153 in ( [ vor ] ) , we imply that @xmath154 and neglect , for simplicity , the effect caused by a dependence of @xmath152 on the polar angle . since @xmath155
, the phase gradient @xmath156 should be nonzero to compensate the drag effect . in the polar
coordinates the @xmath134 component of the phase gradient is given by the relation @xmath157 where @xmath158 the quantity @xmath159 yields the ratio between the phase gradients in the drag and the drive components in the situation when the drag component is in the open circuit ( i.e. the current can not flow in the circuit ) .
we call this quantity the drag factor .
we imply that @xmath151 and @xmath152 are much larger than the healing lengths that allows to describe the drag effect in quasi - uniform approximation . for definiteness
, we specify the case of @xmath160 and @xmath161 in the overlapping region . in this case
one can neglect the space dependence the drag factor ( see eqs .
( [ 1002 ] ) ) .
at nonzero josephson coupling the current @xmath138 can be nonzero , but it can not exceed the maximum josephson current @xmath162 .
relation ( [ 1004 ] ) remains approximately correct at nonzero josephson coupling , if an inequality @xmath163 is satisfied .
here we specify just such a case .
it is important to emphasize that we consider the situation , when there is only the external josephson effect between two half - torus traps , and there is no internal josephson effect between the drag and the drive species .
the drag force can be considered as an effective vector potential @xmath164 ( in units of @xmath165 ) that corresponds to an effective magnetic flux @xmath166 .
thus , our bose system is similar to the cooper pair box system that implements the josephson charge qubit with the josephson coupling controlled by an external magnetic flux @xcite . to extend this analogy we formulate the model of the bose - einstein qubit subjected by the drag force . in what follows we use the approach of ref .
@xcite . in the two mode approximation
the bose field operators for the drag component can be presented in the form : @xmath167 where @xmath168 and @xmath169 are the operators of creation and annihilation of bosons in the condensates confined in the left(right ) half - torus , and @xmath170 , @xmath171 are two almost orthogonal local mode functions @xmath172 that describe the condensate in the left and right traps @xcite . substituting ( [ 1006 ] ) into hamiltonian ( [ 1 ] ) , we obtain the following expression for the parts of the hamiltonian that depends on the operators @xmath173 and @xmath174 : @xmath175 with @xmath176\psi_{l},\ ] ] @xmath177 @xmath178.\ ] ] the functions @xmath170 and @xmath171 contain the phase factors @xmath179 and @xmath180 , where the phases satisfy eq .
( [ 1004 ] ) . taking these factors into account
, one can choose the following basis for the one mode functions @xmath181,\ ] ] where @xmath182,@xmath183 are the polar angles counted from the centers of @xmath184 and @xmath112 half - torus , correspondingly ( see fig . [ fig3 ] ) .
the angles @xmath185 , defined as shown in fig .
[ fig3 ] , satisfy the relation @xmath186 where @xmath187 and @xmath188 are the radius - vectors of josephson links .
left ( l ) and right(r ) half - torus of the drag component , separated by josephson links a and b.,height=226 ] substituting ( [ 1007 ] ) into eq .
( [ 6f ] ) , using eq .
( [ fff ] ) and taking into account that the functions @xmath170 and @xmath171 overlap in a small vicinity of a and b links , we obtain the following expression for the josephson coupling parameter : @xmath189 where @xmath190 is the `` flux quantum '' and @xmath191.\ ] ] here @xmath192 and @xmath193 are the areas of overlapping of two one mode functions at links @xmath4 and @xmath194 , correspondingly . considering the hilbert space in which the total number operator @xmath195 is a conservative quantity ( @xmath196 ) we present the hamiltonian ( [ 3f ] ) in the following form @xmath197 where @xmath198 is the number difference operator , @xmath199 is the interaction energy , and the quantity @xmath200\ ] ] describes an asymmetry of l and r half - tore . in what follows we imply that the system is in the fock regime @xcite ( @xmath201 ) and use the number representation @xmath202 in this representation the first term in ( [ 1013 ] ) is diagonal . the second term in ( [ 1013 ] ) can be considered as small nondiagonal correction .
but if @xmath203 is biased near one of degeneracy points @xmath204 ( where @xmath205 is integer and @xmath206 ) , the second term in ( [ 1013 ] ) results in a strong mixing of two lowest states ( @xmath207 and @xmath208 ) and the low energy dynamics of the system can be described by a pseudospin hamiltonian @xmath209 where @xmath210 are the pauli operators , and @xmath211 are the components of the pseudomagnetic field . in experiments one can control the parameters @xmath203 , @xmath212 and @xmath213 independently and , consequently , the pseudomagnetic field @xmath214 can be switched arbitrary .
it mean that eq .
( [ 1019 ] ) represents the standard hamiltonian of the qubit system .
the parameters of the qubit ( [ 1019 ] ) depend on the `` drag flux '' @xmath215 .
therefore , one can determine its value from the measurement of the state of the system after a controlled evolution of a certain reproducible initial state .
let us consider two possibilities . for definiteness , we specify the case of odd @xmath216 and the degeneracy point @xmath217
if the josephson coupling are switched off and @xmath203 is switched on to some positive value ( much less than unity ) the system is relaxed to the state @xmath218 .
this state can be used as the reproducible initial state .
the quantity should be measured is the expectation value of the number difference operator .
in the initial state the expectation value of this operator is @xmath219 when the system is switched suddenly to the degeneracy point @xmath220 and the josephson couplings are switched on for some time @xmath221 the initial state evolves to another state with another @xmath222 .
if one sets @xmath223 the result of evolution ( @xmath224 ) is described by the unitary operator @xmath225 where @xmath226 .
one can see that at time of evolition @xmath227 the expectation value of the number difference operation will be equal to zero . for the case @xmath228 and @xmath229 the operator of evolution reads as @xmath230 with @xmath231 . respectively
, the expectation value @xmath232 will be equal to zero at @xmath233 the ratio @xmath234 depends only on @xmath215 and the quantity @xmath215 can be extracted from the measurements of @xmath235 and @xmath236 .
it is important to note that to provide this scheme one should control only the ratio of @xmath212 and @xmath213 , but not their absolute values .
another possibility can be based on detection of the berry phase @xcite .
( [ 1019 ] ) contains all three components of the field @xmath237 and they can be controlled independently .
the general scheme of detection of the berry phase in such a situation was proposed @xcite .
a concrete realization of this scheme in the josephson charge qubit was described in @xcite . here
we extend the ideas of @xcite to the case of the `` dragged '' bose - einstein qubit .
we start from the same initial state and switch to @xmath223 and @xmath220 .
the initial state @xmath238 can be presented as the superposition of two instantaneous eigenstates @xmath239 and
@xmath240 : @xmath241 ) results in appearance of the berry phase in the @xmath242 and @xmath243 eigenstates , if the vector @xmath244 subtends a nonzero solid angle at the origin .
let us consider the following 4 stage cyclic adiabatic evolution starting from the point @xmath223 and @xmath220 : 1 - @xmath213 is switched off ; 2 - @xmath212 is switched off and simultaneously @xmath245 is switched to @xmath246 ; 3 - @xmath203 is returned to the same degeneracy point ( @xmath220 ) and @xmath213 is switched to @xmath247 ; 4 - @xmath212 is switched to @xmath228 ( all switches should be done slowly : @xmath248 ) .
after such an evolution the system arrives at the state @xmath249 where @xmath250 is the berry phase ( equals to half of the solid angle subtended by @xmath237 ) and @xmath251 , @xmath252 are the dynamical phases .
elimination of the dynamical phases can performed by swapping the eigenstates ( @xmath253-transformation ) and repeating the same cycle of evolution in a reverse direction ( see @xcite ) .
the @xmath253-transformation can be done by fast switching off the josephson coupling and switching on @xmath254 during the time interval @xmath255 .
after the @xmath253-transformation the state becomes @xmath256 after the cyclic evolution in the reverse direction we arrive to the state @xmath257 one can see that the expectation value of the number difference operator in the final state ( [ 304 ] ) @xmath258 depends only on @xmath215 and the measurement of this difference allows the determine the value of the `` drag flux '' .
thus , the measurements of relative number of atoms in left and right condensates under controlled evolution of the state of the system allows to observe the non - dissipative drag and determine the drag factor ( if the vorticity of the drive component is known ) .
we have investigated the non - dissipative drag effect in three - dimensional weakly interacting two - component superfluid bose gases .
the expression for the drag current is derived microscopically for the general case of two species of different densities , different masses and different interaction parameters .
it is shown that the drag current is proportional to the square root of the gas parameter .
the drag effect is maximal at zero temperatures and it decreases when the temperature increases , but at temperatures of order of the interaction energy the drag current remains of the same order as at zero temperature . we have considered the toroidal double - well geometry , where the non - dissipative drag influences significantly on the josephson coupling between the wells . in the system considered the drag force can be interpreted as an effective vector potential applied to the drag component .
the effective vector potential is equal to @xmath259 ( in units of @xmath165 ) , where @xmath260 is the phase of the drive component , and @xmath159 is the drag factor . in the toroidal geometry
the effective vector potential can be associated with an effective flux of external field @xmath261 , where @xmath262 is the vorticity of the drive component . in the fock regime
the system can be considered as a bose - einstein counterpart of the josephson charge qubit in an external magnetic field .
the measurement of the state of such a qubit allows to observe the drag effect and determine the drag factor .
this work is supported by the intas grant no 01 - 2344 . 99 a. f. andreev and e. p.bashkin , _ zh . exp .
fiz . _ * 69 * , 319 ( 1975 ) [ sov .
jetp , * 42 * , 164 ( 1975 ) ] .
a. j. leggett , rev .
phys . * 47 * , 331 ( 1975 ) .
alpar , s.a .
langer , and j. a. sauls , astrophys .
j. * 282 * , 533 ( 1984 ) .
e.babaev , phys.rev .
d * 70 * , 043001 ( 2004 ) . j. m. duan and s. yip , _ phys
_ * 70 * , 3647 ( 1993 ) .
b. tanatar and a.k .
das , _ phys .
_ b * 54 * , 13827 ( 1996 ) .
s. v. terentjev and s. i. shevchenko , _ fiz .
_ , * 25 * 664 ( 1999 ) [ _ low temp .
phys . _ * 25 * 493 , ( 1999 ) ] .
d.v.fil , s.i.shevchenko , _ fiz .
_ , * 30 * 1028 ( 2004 ) [ _ low temp .
phys . _ * 30 * 770 , ( 2004 ) ] .
leggett , _ rev .
phys . _ * 73 * , 307 ( 2001 ) .
s. i. shevchenko , _ fiz .
temp . _ * 18 * 328 , ( 1992 ) [ _ sov . j. low temp .
phys . _ * 18 * 223 , ( 1992 ) ] .
d. v. fil and s. i. shevchenko , _ phys . rev . _ a * 64 * , 013607 ( 2001 ) .
a.s.parkins , d.f .
walls , _ phys .
rep . _ * 303 * , 1 ( 1998 ) .
n. n. bogolyubov and n. n. bogolyubov ( jr ) , _ introduction to quantum statistical mechanics _ , singapore : world scientific , ( 1982 ) .
s. stringari , _ phys .
lett . _ * 77 * , 2360 ( 1996 ) .
p. hberg , e.l .
surkov , i. tittonen , s. stenholm , m. wilkens and g.v .
shlyapnikov , _ phys .
rev . _ a * 56 * , r3346 ( 1997 ) .
khawaja , c. j. pethick and h. smith , _ phys .
rev . _ a * 60 * , 1507 ( 1999 ) .
m.v.demin , yu . e. lozovik , v.a.sharapov , _
pisma v zhetf _ , * 76 * , 166 ( 2002 ) [ _ jetp letters _ , * 76 * , 135 ( 2002 ) ] z. b. chen and y. d. zhang , _ phys .
rev . _ a * 65 * , 022318 ( 2002 ) .
y. makhin , g. schon and a. shnirman , _ rev .
phys . _ * 73 * , 357 ( 2001 ) .
s. raghavan , a. smerzi , s. fantoni and s.r .
shenoy , _ phys .
_ a * 59 * , 620 ( 1999 )
. m. v. berry , _ proc .
soc . _ a * 392 * , 47 ( 1984 ) .
a. ekert , m. ericsson , p. hayden , h. inamori , j. a. jones , d. k. l. oi and v. vedral , _
_ * 47 * , 2501 ( 2000 ) .
g. falci , r. fazio , g. m. palma , j. siewert and v. vedral , _
nature_(london ) * 407 * , 355 ( 2000 ) . | a microscopic theory of a non - dissipative drag in a two - component superfluid bose gas is developed .
the expression for the drag current in the system with the components of different atomic masses , densities and scattering lengths is derived .
it is shown that the drag current is proportional to the square root of the gas parameter .
the temperature dependence of the drag current is studied and it is shown that at temperature of order or smaller than the interaction energy the temperature reduction of the drag current is rather small . a possible way of measuring the drag factor is proposed . a toroidal system with the drag component confined in two half - ring wells separated by two josephson barriers is considered . under certain condition
such a system can be treated as a bose - einstein counterpart of the josephson charge qubit in an external magnetic field .
it is shown that the measurement of the difference of number of atoms in two wells under a controlled evolution of the state of the qubit allows to determine the drag factor . |
in the present appendix we will describe the procedure to obtain the exact result for the acceleration caused by a distribution of matter in a infinitely thin ring going from @xmath27 ua to @xmath31 ua with a surface density distribution : @xmath62 .
the method is straightforward , using newton s law of gravitation , we calculate the acceleration caused by the distribution of matter on a test particle ; the case of interest for us , is only when the position of the test particle ( the pioneers ) is on the plane of the disk , therefore we need only two dimensions to do the calculations . where @xmath67 is the surface density of the mass element .
then all we have to do is to sum the contribution of all the mass elements in the ring .
to do so , we use the following simplifications : if we assume that @xmath68 , then the symmetry of the problem allow us to put , choosing a cartesian coordinate system , @xmath69 and @xmath70 , where @xmath71 and @xmath72 are the coordinates of the mass element @xmath63 in polar coordinates . the total contribution to
the acceleration in the direction of @xmath73 is zero given the symmetry of the ring ; for the component in the direction @xmath74 the integral over the whole distribution ( @xmath75 going from @xmath76 to @xmath77 and @xmath71 going from @xmath33 to @xmath78 ) gives the radial acceleration caused on the test particle as a function of it s radial position . to perform these calculations we made a numerical program in fortran 77 .
the resulting acceleration as a function of the position @xmath22 of the test particle , appears in fig .
( [ fig : kuiper ] ) of the paper , the value of @xmath79 was adjusted to obtain the magnitude of the acceleration in the pioneer anomaly , the value is : [ alfa_s ] _
s=2.4310 ^ 14 99 m. m. nieto , s. g. turyshev , j. d. anderson , gr - qc/0411077 .
m. m. nieto , s. g. turyshev , j. d. anderson , gr - qc/0501626 v2 .
m. m. nieto , s. g. turyshev , j. d. anderson , physics/0502123 .
mbelek , gr - qc/0407023 .
o. bertolami , j. paramos , _ class .
* 21 * , 3309 , ( 2004 ) .
h. quevedo , gr - qc/0501006 .
r. anania , m. makoid astro - ph/0502582 .
ranada , _ foundations of physics _ , * 34 * , 1955 ( 2004 ) .
e. peebles , and b. ratra _ rev .
phys . _ * 75 * p. 9 , ( 2003 ) .
wmap collaboration ( d.n .
spergel et al . ) , _ apj suppl . _
* 148 * 175 , ( 2003 ) .
l. smolin s remark at a conference during the _ sixth school of the gravitation and mathematical physics division _ ,
playa del carmen , mxico , ( 2004 ) .
b. gladman , _ science _
, * 307 * , 71 , ( 2005 ) .
a. trujillo , m. e. brown , _ apj _ , * 554 * , l95 , ( 2001 ) .
m. m. nieto , astro - ph/0506281 .
j. d. anderson , p. a. laing , e. l. lau , a. s. liu , m. m. nieto , and s. g. turyshev , _ phys.rev . _ * d65 * , 082004 , ( 2002 ) .
gr - qc/0104064 navarro , j. f. , & frenk , c. s. , & white , s. d. m. 1997 , _ apj _ , * 490 * , 493 , e - print : astro - ph/9611107 .
feng , j.l .
, 2004 , annals phys .
315:2 - 51 .
feng , j.l .
, 2005 , submmited to j. phys .
g , astro - ph/0511043 .
bernabei et al .
[ dama collaboration ] , 2000 , phys .
majorovits , et al . , 2004 , astro - ph/0411396 , to be published in the proceedings of the 5th international workshop on the identification and detection of dark matter idm 2004 , edinburgh , sept . , 2004 , world scientific .
n. n. gorkavyi , l.m .
ozernoy , and t. taidakova , e - print : astro - ph/9812480 , ( 1998 ) .
o. bertolami , p. vieira , astro - ph/0506330 | in this work we study the gravitational influence of the material extending from uranus orbit to the kuiper belt and beyond on objects moving within these regions .
we conclude that a density distribution given by @xmath0 ( for @xmath1 ) generates a constant acceleration towards the sun on those objects , which , with the proper amount of mass , accounts for the blue shift detected on the pioneers space crafts .
we also discuss the effect of this gravitational pull on neptune , and comment on the possible origin of such a matter distribution .
mankind is now in direct contact with regions beyond the solar system .
the space probes launched in the @xmath2 already passed the orbit of the last solar system planet , pluto , and are still working !
it is really a homage for their designers .
in particular , the pioneer probes 10 and 11 were designed in such a cunning way that their position , via the doppler effect , can be determined with great accuracy ( the tracking system have the sensitivity to measure frequency changes at the level of @xmath3/s , @xcite ) . when these probes were still within the solar system , at 20 ua , that is , between uranus , which has a mean distance from the sun of 19.13 ua , and neptune , at 30 ua , the frequency received at earth start showing an unaccounted effect , a blue shift that is usually interpreted as a constant acceleration with a magnitude of a=8.741.3310 ^ -8 , [ pioneer]and directed towards the sun @xcite .
the cause for such a blue shift was unknown but more remarkable was the fact that such phenomenon kept being present . in figure 1 , we show a diagram of the solar system showing the pioneers trajectories and positions .
this figure is a reproduction taken from the original at @xcite .
the pioneers trajectories in the solar system @xcite.,width=264 ] there have been a number of possible explanations for the cause of the blue shift , and there is now an important number of works that discard any internal effect of the probes as the cause , such as heat reflected off the probes and possible gas leaks , ( see @xcite , @xcite and @xcite ) , implying that it is due to an external cause .
now the pioneers are farther than @xmath4 ua , well beyond pluto s orbit , @xmath5 ua , and the effect is still there .
following occam s razor , the simplest explanation is that a * constant * force , independent of the distance , is producing the measured blue shift . in this way , the facts are that our first encounter with the external regions of the solar system and beyond are showing the presence of something out there which affects the motion of bodies .
it is interesting to mention that this effect started clearly as the probes passed uranus orbit , still within the solar system , and it has not being observed , so far , in the trajectories of the planets in that region , neptune and pluto .
any number of explanations have being put forward to account for this effect , including dark energy ( see for example @xcite ) , quantum oscillations of the spacetime , branes ( @xcite ) and you name it .
( for other interesting alternatives see @xcite , @xcite and @xcite ) .
our point of view is try first to explain the phenomenon with local , everyday physics , and if this is not enough , then use other alternatives . also , it is clear that dark energy contribution to the motion of bodies at local scales is very much smaller than the one detected at the pioneers trajectories .
indeed , the effects of dark energy start to being noticeable only at galaxy cluster scales ! ( see , for instance , @xcite ) .
however , it is remarkable that , taking the accepted value for the cosmological constant , considering it as the source of the dark energy , @xmath6 , with @xmath7 , the ratio of dark energy density to the critical density of the universe , @xmath8 the hubble constant today ( we take @xmath9 @xcite ) , and @xmath10 the speed of light , we get @xmath11 , and if we want to construct an acceleration associated with it , we obtain that @xmath12 , a value close to the one observed on the pioneers @xcite .
nevertheless , as mentioned , the cosmological effects are negligible at solar systems scales , so this is just a remarkable coincidence . in the present work we develop a more local and common idea .
the solar system started from the proto - planetary cloud which , by gravitational collapse , formed the sun and planets but some material remained in the form of small structures , tiny rocks , and dust , revolving around the sun , forming belts .
those tiny rocks within the orbits of the planets , were ultimately swept by them , and the ones beyond form belts , two mayor regions beyond neptune s orbit : the kuiper belt , going from @xmath13 ua to @xmath4 ua , where it joints with the ort cloud , which extends up to 4000 ua @xcite , and internal belts within the orbits of the four major external planets .
we study the gravitational pull generated by those belts on the objects moving within them and conclude that , if their density distribution goes as @xmath14 , a constant acceleration pull is produced , directed towards the center , which could account for the observed blue shift on the pioneer probes . as an illustration ,
let s take the simplest case of a solid spherical distribution of matter , an object moving within a media of a given density distribution @xmath15 , will be accelerated according to the following newtonian law : = -4 g ( ) , with @xmath16 the gravitational constant , @xmath17 , and @xmath18 the acceleration of the object .
thus , _
sda=-4g_v()d^3r , with @xmath19 the normal vector to the surrounding surface to the volume @xmath20 .
considering that the acceleration is only radial : @xmath21 , and taking a spherical shell of radius @xmath22 , as the surface of integration , we get a_p=-4g_0^r(r)r^2dr , [ eq : a_p]from which is clear that , in order to have a constant acceleration , the density on the media must go as : ( r)=[eq : rho ] moreover , we can compute the proportionality constant @xmath23 , from the observed value of the pioneers acceleration ; obtaining that : = = 0.21 and then , the density of the sphere needed to produce such an acceleration is , for example at @xmath24 ua where it s claimed that the anomaly starts : @xmath25 . however , this solid sphere is not the answer to the problem because it starts form @xmath26 , i.e. from the sun , and the anomaly does nt appear until @xmath24 ua . the next possibility is a spherical shell starting at @xmath24 ua with the same density distribution as in eq .
( [ eq : rho ] ) . in this case
the acceleration within the shell is not constant : for the region inside @xmath24 ua the acceleration is zero , for the region inside the shell , we can calculate the acceleration by taking the difference between the influence of two spheres centered in the sun ; the smaller one with a radius @xmath27 ua and a mass @xmath28 produces a radial acceleration in a test particle located at @xmath29 given by : a_1=-=- the bigger sphere with radius @xmath22 produces an acceleration whose magnitude is given by eq .
( [ eq : a_p ] ) : @xmath30 ; therefore , the acceleration inside the shell for a test particle is : a_p=-2g(1- ) .
deacceleration caused in a test particle located at @xmath22 due to a spherical shell starting from @xmath27 ua to @xmath31 ua , the acceleration is in units of @xmath32.,width=188 ] for this case the resulting acceleration is not independent of the radius but as @xmath22 gets larger compared to @xmath33 the acceleration approaches a constant value .
if we fix this value to eq .
( [ pioneer ] ) , then the value of @xmath23 and @xmath34 are the same as in the case of the solid sphere . even when the acceleration is not constant within the sphere , this case ca nt be discarded because of the uncertainty in the value of the anomaly ( around @xmath35 , see eq .
( [ pioneer ] ) ; see also fig . 2 of @xcite ) ; from figure [ fig : shell ] we can see that for an important region inside the sphere the calculated acceleration is in accordance with the anomaly taking into account the uncertainty in the observational data .
the third case is a cylindrical ring , with height @xmath36 and borders at @xmath27 ua and @xmath31 ua and a surface density @xmath37 ( in this case , @xmath22 is the radius in cylindrical coordinates ) , and the solution is not as trivial as the two previous . in the appendix
we describe the method used to calculate the acceleration of an infinitely thin disk ( @xmath38 ) , the result appears in figure ( [ fig : disk ] ) .
the following important features can be seen in the figure : the acceleration is almost constant between @xmath39 ua and @xmath40 ua with a change of less than @xmath41 ; near the edges of the ring a significant change in the acceleration takes place , this is due to the fact that @xmath18 is singular at the edges . acceleration caused in a test particle located at @xmath22 due to an infinitely thin disk going from @xmath27 ua to @xmath31 ua
, the acceleration is in units of @xmath42.,width=264 ] the first feature of the results show us that a constant acceleration can be produced in a large region inside the ring by such a simple model .
the region of this constancy can be extended backwards ( from @xmath39 ua to @xmath24 ua as the pioneer anomaly requires ) by extending backwards the ring , it has to start then at @xmath43 ua .
however it s important to mention that in order to give a better estimate of where this ring could start in reality , it s necessary to have more precise data that the one reported for the pioneer anomaly , specially in the zone when the anomaly starts to be significant , because it s in this region where the larger uncertainty in the data is present ( see figure 2 of @xcite ) . regarding the second feature
, we can say that the singularities can disappear taking a smooth but fast change in the surface density in the edges and not a cut - off as the one we used in the calculations , this is is of course a more realistic case .
it s also important to say at this point that when the 2d case we have chosen here is taken to 3d for a a cylindrical ring of height @xmath36 the results does nt change significantly , the calculation has been made somewhere else @xcite ( see figures 5 and 8 of the paper ) , for this case , in the region of constant acceleration , the change in magnitude between the 2d and 3d models is almost null , the main difference is near the edges where the spikes become less extreme for the 3d case . in order to obtain the magnitude we need for the acceleration , the surface density at @xmath44 ua of the thin ring must be , according to eq .
( [ eq : a_p ] ) in the appendix : ( 20ua)=0.81 in the realistic case of 3d this ring becomes a cylindrical ring of height @xmath36 that has a volumetric density given by @xmath45 , then for @xmath44 ua , and assuming @xmath46 ua we have : @xmath47 .
the total mass of a belt with such a density , considering a disk with a thickness of @xmath48 ua , and ranging from @xmath24 to @xmath49 ua , gives * @xmath50 * , two order of magnitudes larger than the current estimates on the kuiper s belt mass @xcite .
a first possible explanation for this discrepancy , which is obtained considering that the belt is formed out of pure dust , could be the presence of tiny ice rocks and gas which have not been accounted for by those studies . in the same work , @xcite
, the authors did consider , as in our present work , the gravitational influence of the kuiper belt as a possible explanation to the acceleration seen on the pioneers , but discarded that influence based on two reasons : first that the acceleration profile is not constant across the data range , ( this is because they did not consider a density distribution going as @xmath51 ) , and second that the total mass exceeded the current estimations , @xmath52 , for the dust in the kuiper belt region from @xmath13 to @xmath49 ua . nevertheless , even though such mass distribution does explain the deacceleration observed in the pioneers , it is unlikely that baryonic matter could be responsible for this effect , there would have to be a very high amount of small objects implying a large collision rate and optical and infrared signatures that have not been detected .
thus , in order to keep the model and mantain that the observed deacceleration is due to gravitational pull , the other possible candidate is the dark matter within the solar system .
the fact that the galaxies , including ours of course , are surrounded by dark matter halos is already a well established fact .
the numerical simulations , namely the nfw , predicts the dark matter halo and give a well known density distribution @xcite . evaluating that density at the position of our solar system , and considering that the gravitational influence of the sun extends to the ort cloud ( around @xmath53 ua )
, we see that there are about @xmath54 of dark matter whose dynamics is dictated by the sun .
it is still an open question on how this solar dark matter halo reacts to that influence . when the proto solar system was being formed , the solar dark matter halo was pulled inwards and it could be that the final configuration is an spherical shell or a belt in a region close to the sun with properties similar to the ones described in the last paragraphs , giving in this way the mass needed to explain the observed deacceleration on both spacecrafts . * given the properties requiered for the dark matter particles , the natural candidates are wimps ( weakly interacting massive particles ) . indeed ,
one most favoured wimp to be the dark matter particle is the neutralino , predicted by one of the extensions of the standard model of particles , the minimal supersymmetryic standard model ( mssm ) , which complies with the characteristics needed for a dark matter particle , namely mass of the order of @xmath55 , making it a member of the cold dark matter pradigm , very small cross section , @xmath56 , electrically neutral and stable ( @xcite ) .
detection of the neutralino is nowadays an exciting field of research , neutralinos may be detected either directly through their interactions with ordinary matter or indirectly through their annihilation decay products . in the first case
, several experiments on earth are looking for signals associated with dark matter scattering with baryonic matter ( nucleus recoils for example ) ( @xcite ) with no conclusive result yet .
the interpretation of this data depends on the spatial distribution of density and velocity for dark mater in the earth neighborhood which is poorly known .
the model presented here could serve as an attempt to give a prediction for the density distribution of dark matter using the data of the pioneer s anomaly , but only for regions beyond @xmath57 , therefore it has no effect for direct detection experiments on earth .
indirect detection is based on looking for signals of the remnants resulting from dark matter annihilation , the neutralino being a majorana particle can annihilate with itself .
the production of these remnants , positrons , photons , neutrinos , etc .
, can be enhanced in regions of high dark matter density such as the center of galaxies ( for an excellent review on this topic see @xcite ) . * * the properties of the neutralino fit very well with the model presented in this work .
their presence influence the dynamics of the objects mostly through gravitational interaction and has a negligible dispersion with ordinary matter . and
it is for this last reason that there could not be a measurable contribution to the pionner anomaly due to dispersion effects . *
also , it is important to point out that the belt would also affect neptune s orbit , modifying its period in the following way ( using the approximation of circular orbits ) : t^2=4 ^ 2 ( ) , that is , neptune also feels the acceleration due to this matter distribution .
such acceleration changes neptune s period in @xmath58 seconds _ per _ period , implying a shift in neptune s center of mass of @xmath59 kilometers after each revolution .
it is a very tiny quantity but future probes could be designed in such a way as to be able to measure these effects .
the radial density profile ( going as @xmath60 ) needed to explain a constant acceleration towards the sun , can be explained by models of solar system formation ( in the case of baryonic matter only ) . in @xcite
, the authors made an analysis of the dust population in the outer solar system region ; making a computational approach ( see fig .
[ fig : kuiper ] ) considering source objects from the jupiter - family comets and the kuiper belt .
they obtained a radial distribution of dust for the region between 20 to 70 ua , that has a sharp peak near 20 ua and from there falls like @xmath51 , just as our model predicts .
after that , it has a smaller peak near 50 ua ( see fig .
[ fig : densidad ] of this paper that is a reproduction of fig .
5 of @xcite ) .
this result is encouraging for our model but given the large amount of mass needed to reproduce the actual deacceleration measured by the pioneers , it would have to be remade with the inclusion of dark matter to see if the density profile remains the same with the required amount of mass .
numerical simulation of the density distribution of matter at the solar system in the ecliptic plane @xcite.,width=264 ] numerical simulation of the radial density distribution of matter at the solar system in the ecliptic plane @xcite.,width=302 ] following this scheme , the mass density distribution that appears after the orbit of uranus , is a result of an inward transport of material from the kuiper belt into this orbit due mainly to the gravitational influence of neptune and uranus . if the radial distribution goes in fact like @xmath51 ( at least between 20 to 100 ua ) , as the kind of models presented by gorkavyi _
et al . _
@xcite , and with a total mass of @xmath61 , then the so called pioneer anomaly can be explained by the gravitational force that this distribution produce on every test particle inside it , as was shown previously , making the observed blue shift not an anomaly , but a natural consequence of the gravitational pull due to the material present in these regions . * indeed , the actual dark matter distribution within the solar system is still an open question .
a numerical simulation has to be performed to give an appropriate description on this subject , taking into account the effects of the formation and evolution of the solar system on the dark matter solar halo .
the outcome of these analysis would determine whether or not the matter distribution proposed in the present model is plausible .
in particular , it is important to determine the influence due to the outer planets .
this would answer the question if a gap of @xmath24 au starting from the sun an then a region with density falling like @xmath60 that has not been trashed away by neptune is possible .
the gravitational influence of the outer planets on the baryonic matter has formed solar belts in these regions , which probably has a matter distribution similar to the one required for dark matter . *
also , we want to comment that while this work was being finished , we learned that nieto _ et .
al _
@xcite were also considering the effect of the material in the kuiper belt on the space probes .
however , in their approach they studied the drag force of this material on the probes , obtaining a different density distribution , mainly that the density should be constant in order to explain the constant acceleration pull observed on the pioneers . finally , we remark that it is very important to perform new and more detailed studies on the density distribution and total mass of the matter belts in the regions beyond uranus orbit to verify , or discard , the explanation for the pioneer acceleration presented in this work . * in order to do so , we consider it is necessary to send several probes to the outer regions of the solar system , like the tau probe ( thousand astronomical unit ) , in several directions , including those perpendicular to the solar system plane , and equipped with the needed instruments in order to determine , via the deacceleration effects , the nature and characteristics of the phenomenon which causes these effects , and then see if it is compatible with a dark matter distribution . * we acknowledge that the required amount of mass needed to reproduce the magnitude of the pioneers deacceleration is indeed very large considering the actual estimates of mass in such a region , however the idea proposed in this work is appealing due to it s simplicity and has already generated some attention in the scientific community ( see for example the works that have appeared while this work was being considered for publication , @xcite ) .
we think that the proposal described above considering dark matter as a possible explanation for the unaccounted mass is worth to follow and , as we mentioned above , it would be interesting to perform numerical simulations for the solar system formation which include dark matter in order to study its final distribution .
this work was partially supported by the grants dgapa - unam in113002 , in122002 .
jz also acknowledges the scholarship support by conacyt and dgep .
we want to thank our colleagues in the gravitational department of the nuclear science institute for much encouragement and helpful discussions on the present idea .
we also thank the anonymous referee of the present work for several remarks which contributed to better sketch the problems and advantages of the model proposed . |
we continue our development on algorithms to count the points of varieties related to quiver representations . in this note ,
we focus on the quivers with potentials .
a _ potential _
@xmath0 on a quiver @xmath1 is just a linear combination of oriented cycles of @xmath1 .
it can be viewed as a function on certain noncommutative space attached to @xmath1 .
when composing with the usual trace function , it becomes a regular function @xmath2 on each representation spaces @xmath3 .
this function further descends to various moduli spaces .
let @xmath4 be a regular function on a complex variety @xmath5 .
consider the scheme theoretic degeneracy locus @xmath6 of @xmath4 .
behrend , bryan and szendri define in @xcite a class @xmath7\in k_0({\operatorname{var}}_{\mb{c}})[\mb{l}^{-\frac{1}{2}}]$ ] associated to each such locus , essentially given by the motivic _ milnor fibre _ of the map @xmath4 .
when @xmath5 admits a suitable torus action , this class can be expressed as$ ] differs from the original one by a negative sign . ]
@xmath8 = [ f^{-1}(0 ) ] - [ f^{-1}(1)].\ ] ] deligne s mixed hodge structure on compactly supported cohomology gives rise to the @xmath9-polynomial homomorphism @xmath10 $ ] given by @xmath11;x , y)=\sum _ { p , q } x^p y^q \sum(-1)^i\dim h_{p , q}(h^i_c(y,\mb{q})).\ ] ] following n. katz ( * ? ? ?
* appendix ) , we say that an element @xmath12 is _ fibrewise polynomial - count _ , if for every finite field @xmath13 , and every ring homomorphism @xmath14 , the element @xmath15 deduced from @xmath16 by extension of scalars is polynomial - count ( but we allow its counting polynomial to vary with @xmath17 ) .
it is known ( ( * ? ? ?
* theorem 6.1.2 , 6.1.4 ) ) that [ l : polycount ] if @xmath12 is fibrewise polynomial - count , then all counting polynomials coincide as one @xmath18 $ ] with @xmath19 . in this notes , we follow an algebraic approach to compute @xmath20 directly in the above quiver setting .
namely , @xmath16 is the class @xmath21 $ ] , defined by the regular function @xmath2 on representation spaces .
in fact , we will work with the generating series of _ virtual class _
@xmath22 for all dimension vectors : @xmath23 our main results contain two wall - crossing formula one for the ordinary @xmath24 , the other for the one with a _ framing stability _
@xmath25 : @xmath26 where @xmath27 denotes the git moduli of @xmath28-dimensional @xmath29-stable representations . to state them ,
let @xmath30 be the mutation of @xmath31 in the sense of derksen , weyman , and zelevinsky @xcite .
let @xmath32 and @xmath33 , then under the technique condition @xmath34 of section [ s : qpmu ] , we have that @xmath35 here , multiplications are performed in appropriate completed quantum laurent series algebras .
@xmath36 and @xmath37 are certain linear monomial change of variables .
there are two main ingredients in the proof .
one is a construction of mozgovoy , which relates the hall algebra of a quiver to the quantum laurent series ring .
the other is a ` dimension reduction ' technique used by a. morrison and k. nagao .
the equation already appeared in @xcite and @xcite , but we put it in a right assumption .
nagao also suggested in @xcite that this theory can be used to study quantum cluster algebra .
we follow his suggestion and use to categorify quantum cluster algebras under the assumption of existence of certain potentials .
if a cluster algebra has such a categorification , then its _ strong positivity _ will be implied by a result of @xcite on the purity of the vanishing cycles .
let @xmath38 and @xmath39 be two skew - symmetric matrices of size @xmath40 and @xmath41 @xmath42 .
we extend @xmath39 from the right to an @xmath43 matrix @xmath44 .
we assume @xmath38 and @xmath45 are unitally compatible , that is , @xmath46 $ ] .
we can associate @xmath44 a quiver @xmath1 without loops and 2-cycles with @xmath47 we endow @xmath1 with some potential @xmath0 of nice properties .
let @xmath48 be a sequence of _ mutations _ , and set @xmath49 .
let @xmath50 be some initial _ cluster monomial _ in the quantum laurent polynomial ring @xmath51 .
we extend qp @xmath52 from @xmath53 by adding a new vertex @xmath54 and @xmath55 new arrows from @xmath56 to @xmath54 .
we apply the inverse of @xmath48 to @xmath52 , and obtain a qp @xmath57 .
let @xmath58 be the extended @xmath44-matrix of @xmath59 .
the _ mutated _ cluster monomial @xmath60 is equal to @xmath61 this result may just be a special case of @xcite , where a. efimov assumed a much weaker condition on the potential @xmath0 . however , our approach and result are more down - to - earth and computable .
it depends only on @xcite rather than @xcite .
the second application is on the generating series counting the quiver representation grassmannian .
let @xmath62 be a sink of @xmath1 , and @xmath63 be a representation of @xmath1 .
we assume that @xmath63 does not contain the simple representation @xmath64 as a direct summand .
let @xmath65 where @xmath66 is the variety parameterizing @xmath67-dimensional quotient representations of @xmath63 .
@xmath68 and @xmath69 are also related via . in particular , if @xmath63 is _ polynomial - count _ , that is , all its grassmannians @xmath66 are polynomial - count , then so are all reflection equivalent classes of @xmath63 .
this note is organized as follows . in section [ s :
def ] , we recall some basics about quiver representations and their hall algebras . in section [
s : qpc ] , we recall the concept of quiver with potential and a cut , and its associated algebras . in section
[ s : mozgovoy ] , we recall a construction of mozgovoy and the dimension reduction technique ( lemma [ l : vtor ] ) . in section [
s : qpmu ] , we recall the mutation of qp with a cut , and set up the key assumption for our main results .
when this assumption holds is illustrated in corollary [ c : tilting ] , whose proof will be given in the appendix .
in section [ s : wc ] , we prove our first two main results
theorem [ t : wc ] and theorem [ t : fwc ] . in section [
s : cluster ] , we prove our third main result , theorem [ t : cluster ] , on a categorification of quantum cluster algebras . in section [ s : qgrass ] , we prove our last main result , theorem [ t : qgrass ] , on the representation grassmannians under reflections . 1 .
all modules are right modules and all vectors are row vectors .
2 . for an arrow @xmath70
, @xmath71 is the tail of @xmath70 and @xmath72 is the head of @xmath70 .
3 . for any representation @xmath63
, we use @xmath73 to denote its dimension vector .
@xmath74 is the simple module at the vertex @xmath56 , and @xmath75 is its projective cover .
superscript @xmath76 is the trivial dual @xmath77 .
from now on , we assume our base field @xmath78 to be the finite field with @xmath79 elements . let @xmath1 be a finite quiver with the set of vertices @xmath80 and the set of arrows @xmath81 .
we write @xmath82 for the usual euler form of @xmath1 .
we also have the antisymmetric form @xmath83 associated to @xmath1 .
the matrix of @xmath83 denoted by @xmath44 is given by @xmath84 for any three @xmath85-modules @xmath86 and @xmath0 with dimension vector @xmath87 and @xmath88 , the _ hall number _ @xmath89 counts the number of subrepresentations @xmath90 of @xmath0 such that @xmath91 and @xmath92 .
we denote @xmath93 let @xmath94 be the space of all formal ( infinite ) linear combinations of isomorphism classes of @xmath85-modules .
* proposition 1.1 ) the completed _ hall algebra _ @xmath94 is the associative algebra with multiplication @xmath95[v]:=\sum_{[w]}f_{uv}^w[w],\ ] ] and unit @xmath96 $ ] .
let @xmath97 be a subcategory of @xmath98 .
we denote @xmath99 $ ] .
we use the shorthand @xmath100 and @xmath101 for @xmath102 and @xmath103 .
let @xmath104 be a _ torsion pair _
( ( * ? ? ?
* definition vi.1.1 ) ) in @xmath98 , then for any @xmath105 , there exists a short exact sequence @xmath106 with @xmath107 unique in @xmath108 and @xmath109 unique in @xmath110 . in terms of the hall algebra
, this says that [ l : torsionid ] @xmath111 a _ weight _ @xmath112 is an integral linear functional on @xmath113 .
a _ slope function _
@xmath29 is a quotient of two weights @xmath114 with @xmath115 for any non - zero dimension vector @xmath28 .
a representation @xmath63 is called _ @xmath29-semi - stable ( resp .
@xmath29-stable ) _ if @xmath116 ( resp .
@xmath117 ) for every non - trivial subrepresentation @xmath118 .
we denote by @xmath119 the variety of @xmath28-dimensional @xmath29-semistable representations of @xmath1 . by the standard git construction @xcite
, there is a _ categorical quotient _ @xmath120 and its restriction to the stable representations @xmath121 is a _ geometric quotient_. a slope function @xmath29 is called _ coprime _ to @xmath28 if @xmath122 for any @xmath123 .
so if @xmath29 is coprime to @xmath28 , then there is no strictly semistable ( semistable but not stable ) representation of dimension @xmath28 . in this case , @xmath124 must be a geometric quotient . _
* proposition 2.5 ) _ _ harder - narasimhan filtration : _ + every representation @xmath63 has a unique filtration @xmath125 such that @xmath126 is @xmath29-semi - stable and @xmath127 .
we fix a slope function @xmath29 . for a dimension vector @xmath28 ,
let @xmath128 $ ] .
the existence of the harder - narasimhan filtration yields the following identity in the hall algebra @xmath94 .
* proposition 4.8 ) [ l : hnid ] @xmath129 where the sum running over all decomposition @xmath130 of @xmath28 into non - zero dimension vectors such that @xmath131 . in particular , solving recursively for @xmath132 , we get @xmath133 where the sum runs over all decomposition @xmath130 of @xmath28 into non - zero dimension vectors such that @xmath134 for @xmath135 .
* lemma 6.1 ) [ ex : int ] let @xmath136 be the quantum laurent series algebra in @xmath137 variables with multiplication given by @xmath138 .
then @xmath139\mapsto q^{\frac{1}{2}\innerprod{\alpha,\alpha}}\frac{x^\alpha}{a_m}$ ] is an algebra homomorphism @xmath140 . since @xmath141 , we have that @xmath142 .
we fix a quiver @xmath1 without loops and oriented 2-cycles and a potential @xmath0 on @xmath1 . recall that a _ potential _
@xmath0 is an element in @xmath143 $ ] , where @xmath144 is the completion of the path algebra @xmath85 .
when dealing with completed path algebras , all its ideals will be assumed to be closed .
the vector space @xmath143 $ ] has as basis the set of oriented cycles up to cyclic permutation . for each arrow @xmath145 , the _ cyclic derivative _
@xmath146 on @xmath144 is defined to be @xmath147 for each potential @xmath0 , its _ jacobian ideal _
@xmath148 is the ( two - sided ) ideal in @xmath144 generated by all @xmath149 .
let @xmath150 be the _
jacobian algebra_. if @xmath151 is finite dimensional , then the completion is unnecessary .
let @xmath152 be the _ trace function _ corresponding to the potential @xmath0 defined by @xmath153 .
note that @xmath2 is an additive function , so it in fact descends to the grothendieck group : @xmath154 . by abuse of notation , we also use @xmath2 for the same trace function defined on the affine representation varieties and moduli spaces .
the following lemma is well - known . a representation of @xmath1 a representation of @xmath155 if and only if it is in the degeneracy locus of @xmath2 .
following @xcite , we define a _ cut _ of a qp is a subset @xmath156 such that the potential @xmath0 is homogenous of degree @xmath157 for the weight function @xmath158 defined by @xmath159 for @xmath160 and zero otherwise .
note that this weight function defines a @xmath161-action on @xmath3 such that the trace function is equivariant : @xmath162 .
the algebra @xmath163 associated to a qp @xmath31 with a cut @xmath164 is the quotient algebra of the jacobian algebra @xmath155 by the ideal generated by @xmath164 .
note that @xmath163 is isomorphic to degree zero part of @xmath155 .
we denote by @xmath165 the subquiver @xmath166 of @xmath1 and by @xmath167 the ideal @xmath168 .
it is clear that @xmath163 can also be presented as @xmath169 .
we put @xmath170 [ d : exact ] a cut is called _ algebraic _ if 1 .
@xmath163 is a finite dimensional @xmath13-algebra of global dimension @xmath171 , 2 .
@xmath172 is a minimal set of generators of the ideal @xmath167 in @xmath144 .
it is clear that for algebraic cuts , the form @xmath173 is exactly the euler form of @xmath163 . from now on , we assume that all cuts are algebraic .
conversely , given any @xmath13-algebra @xmath174 presented by a quiver @xmath1 with a minimal set of relations @xmath175 , we can associate with it a qp @xmath176 with a cut @xmath164 as follows : @xmath177 with @xmath178 , and @xmath179 .
if @xmath174 has global dimension 2 , it is known ( * ? ? ?
* theorem 6.10 ) that @xmath180 is isomorphic to the algebra @xmath181 . in particular , @xmath180 does not depend on the minimal set of relations that we chose .
it is clear that ( see also ( * ? ? ?
* proposition 3.3 ) ) 1 .
@xmath182 , 2 .
let @xmath164 be an algebraic cut of @xmath31 and @xmath183 , then @xmath31 and @xmath176 are isomorphic qps .
let @xmath5 be a quasiprojective variety with a @xmath161-action , and @xmath2 a regular function on @xmath5 .
we assume that @xmath2 is equivariant with respect to a primitive character .
we set @xmath184 if @xmath5 is a complex variety , and we further assume that @xmath185 exist for all @xmath186 , then according to ( * ? ? ?
* proposition 1.11 ) , @xmath187\in k_0({\operatorname{var}}_{\mb{c}})$ ] is the _ nearby fibre _ of @xmath2 .
its difference with the _ central fibre _
@xmath188 defines a class called the _ ( absolute ) vanishing cycle of _
@xmath2 on @xmath5 : @xmath189:=[\omega^{-1}(0)]-[\omega^{-1}(1)].\ ] ] due to the torus action , we have that @xmath190 we denote the @xmath79-shifted @xmath191 by @xmath192 .
let @xmath193 be the trace function corresponding to the potential @xmath0 .
for @xmath194\in h(q)$ ] , we define @xmath195 $ ] .
such an @xmath196 is called _ equivariant _ if @xmath197 for any @xmath198 .
let @xmath199 be the subalgebra of @xmath94 consisting of equivariant elements .
* proposition 5.2 ) the map @xmath200 defined by @xmath201 is an algebra morphism .
note that if @xmath0 is trivial , then @xmath202 . we see from that @xmath203 we denote the generating series @xmath204 by @xmath205 ( * ? ? ?
* theorem 4.1 ) [ l : vtor ] @xmath206 .
so @xmath207 for any stability @xmath29 , the trace function @xmath2 restricts to @xmath208 and descends to the git moduli space @xmath27 .
note that for @xmath209 , @xmath210-[\omega^{-1}(1)]$ ] is the vanishing cycle of @xmath2 on @xmath27 ( ( * ? ? ?
* theorem 4.3 ) ) .
the only thing we need to verify is that @xmath211 exists in @xmath27 for any @xmath212 .
however , similar statement does not hold for @xmath119 in general .
apply the hall character @xmath213 to the identity , then we obtain ( * ? ? ?
* theorem 5.7 ) : [ c : int1 ] @xmath214 where the summation @xmath76 is the same as in lemma [ l : hnid ] .
we denote by @xmath215 the above rational function in @xmath216 .
following @xcite , we say an algebra @xmath174 is polynomial - count if each @xmath217 is polynomial - count .
assume that the git quotient @xmath27 is geometric .
if @xmath163 is polynomial - count , then so is @xmath218 .
a pair @xmath219 is called _ numb _ to a cut @xmath164 on @xmath1 if the vector bundle @xmath220 restricts to @xmath29-semistable representations .
later we will need the following generalization of lemma [ l : vtor ] .
the proof is the same as that in @xcite . for readers convenience
, we copy the proof here .
[ l : vtormu ] if @xmath29 is numb to @xmath164 , then @xmath221 . so @xmath222 by assumption , @xmath223 is a vector bundle of rank @xmath224 .
the restriction of @xmath2 to the fibre @xmath225 is zero if @xmath226 , and is a non - zero linear function if @xmath227 .
hence @xmath228 by , @xmath229 let @xmath230 be the standard basis of @xmath231 .
@xmath13-th ( absolute ) framing stability _
@xmath30 is the slope function given by @xmath232 , where @xmath233 .
it is not hard to see that if all arrows in @xmath164 end in @xmath13 , then @xmath234 with @xmath235 is numb to @xmath164 .
the key notion in @xcite is the definition of mutation @xmath30 of a quiver with potentials at some vertex @xmath236 .
let us briefly recall it .
the first step is to define the following new quiver with potential @xmath237 .
we put @xmath238 and @xmath239 is the union of three different kinds 1 .
all arrows of @xmath1 not incident to @xmath13 , 2 . a composite arrow @xmath240 $ ] from @xmath71 to @xmath241 for each @xmath70 and @xmath242 with @xmath243 , and 3 . an opposite arrow @xmath244 ( resp .
@xmath245 ) for each incoming arrow @xmath70 ( resp .
outgoing arrow @xmath242 ) at @xmath13 .
the new potential is given by @xmath246+\sum_{ha = tb = k}b^*a^*[ab],\ ] ] where @xmath247 $ ] is obtained by substituting @xmath240 $ ] for each words @xmath248 occurring in @xmath0 .
let @xmath174 be the algebra @xmath163 and @xmath249 be the representation @xmath250 , where @xmath251 .
clearly , @xmath252 can be presented as @xmath253 .
* corollary 2.2.b ) @xmath249 is a tilting module iff .
@xmath254 is injective . in this case
, @xmath249 is called the bb - tilting module at @xmath13 .
the dual notion of @xmath249 is the bb - cotilting module @xmath255 .
what we desire is the following nice situation : + in general , the existence of another cut on @xmath256 is not guaranteed .
however , if we assume that @xmath257 then we can assign a new cut @xmath258 containing all 1 .
@xmath259 if @xmath260 , 2 .
arrows @xmath245 if @xmath261 , 3 . composite arrows
@xmath240 $ ] with @xmath262 .
this definition is related to the graded right mutation defined in @xcite .
there is a graded version of splitting theorem ( ( * ? ? ?
* theorem 4.6 ) ) . applied to qp with a cut ,
we have [ l : gsplit ] @xmath263 is graded right - equivalent to the direct sum @xmath264 , where @xmath265 is _ reduced _ and @xmath266 is _ trivial _ , both unique up to graded right - equivalence .
we denote the reduced part of @xmath267 by @xmath268 .
[ t : tilting ] assume @xmath164 is an cut satisfying definition [ d : exact].(2 ) and , and that @xmath269 for any @xmath270 .
then @xmath163 is tilting equivalent to @xmath271 via @xmath272 .
[ c : tilting ] if @xmath164 is an algebraic cut satisfying , then @xmath273 is also algebraic , and @xmath163 is tilting equivalent to @xmath271 via @xmath272 .
these slightly generalize the main results of @xcite .
we will prove them in the appendix .
by lemma [ l : gsplit ] , the above @xmath271 can be replaced by @xmath274 .
if we want to work with the assumption dual to , that is , all arrows starting with @xmath13 do not belong to @xmath164 , then we should take the functor @xmath275 .
the equivalence @xmath272 induces a map @xmath276 in the corresponding @xmath277-group @xmath278)=\begin{cases } [ s_i ' ] & i\neq k , \\ -[s_k']+\sum_{ha = k}[s_{ta } ' ] & i = k;\end{cases}\end{aligned}\ ] ] and its dual @xmath275 induces @xmath279 given by @xmath280)=\begin{cases } [ s_i ' ] & i\neq k , \\ -[s_k']+\sum_{tb = k}[s_{hb } ' ] & i = k.\end{cases}\end{aligned}\ ] ] by slight abuse of notation , we also write @xmath276 and @xmath279 for the corresponding linear isometries on @xmath231 . due to the equivalence
, we have that @xmath281 .
moreover , it is easy to verify that @xmath282 , or equivalently , @xmath283 , where @xmath284 is the antisymmetric form of @xmath285 .
we denote @xmath286 note that under the assumption @xmath34 , @xmath287 ( resp .
@xmath288 ) is the torsion ( resp .
torsion - free ) class determined by the tilting module @xmath249 .
so @xmath289 in particular , for @xmath290 we have that @xmath291
we denote @xmath292 .
let @xmath293 and @xmath294 .
let @xmath37 ( resp .
@xmath36 ) be the ring homomorphism @xmath295 defined by @xmath296 ( resp .
@xmath297 ) .
[ t : wc ] assuming the condition @xmath34 , we have that @xmath298 we apply the character @xmath213 to the torsion - pair identity in @xmath94 @xmath299 we get @xmath300 the other half is similar .
we freeze a vertex @xmath54 of @xmath1 , that is , we do not allow to mutate at @xmath54 . let @xmath301 be all modules supported outside @xmath54 .
note that @xmath301 is an exact subcategory of @xmath302 .
in particular , it is a torsion - free class , and let @xmath303 be its corresponding torsion class .
let @xmath304 and @xmath305 .
it follows from the torsion pair identity that @xmath306 we keep the assumption @xmath34 . according to theorem [ t : wc ] , we have that @xmath307 we can also treat @xmath308 as a torsion class , and work with its torsion - free class @xmath309 . if we set @xmath310 , then @xmath311 , and we have the dual formula @xmath312 consider the subcategory of @xmath313 of @xmath303 , which contains all representations having dimension one at the vertex @xmath54 .
it is well - known that the class @xmath313 are determined by the framing stability @xmath25 .
let @xmath314 since the dimension vector @xmath315 is coprime to the slope function @xmath25 , the moduli space @xmath316 is a geometric quotient , and thus we have @xmath317 clearly , we also have that [ t : fwc ] @xmath318 in the next section , by abuse of notation , we will write @xmath30 for the operator @xmath319 , and @xmath320 for the operator @xmath321 .
let @xmath38 and @xmath39 be two skew - symmetric matrices of size @xmath40 and @xmath41 @xmath42 .
we extend @xmath39 from the right to an @xmath43 matrix @xmath44 .
we assume @xmath38 and @xmath45 are unitally compatible , that is , @xmath46 $ ] .
we can associate @xmath44 a quiver @xmath1 without loops and 2-cycles satisfying .
the vertices @xmath322 are _ frozen vertices_. we denote by @xmath323 _ the principal part _ of @xmath1 , that is , the subquiver of @xmath1 by forgetting all frozen vertices .
let @xmath51 be the quantum laurent polynomial ring in @xmath324 variables with @xmath325 .
it is contained in its skew - field of fractions @xmath110 .
a toric frame is a map @xmath326 , such that @xmath327 for some automorphism @xmath328 of the skew - field @xmath110 , and some automorphism @xmath329 of the lattice @xmath330 .
let @xmath331 be the standard basis of @xmath330 .
we also denote by @xmath276 the matrix of the linear isometry , and by @xmath332 its restriction on the principal part @xmath323 . for any integer @xmath242 , we write @xmath333_+$ ] for @xmath334 .
a _ seed _ is a triple @xmath335 such that @xmath336 for all @xmath337 .
the _ mutation _ of @xmath338 at @xmath13 is a new triple @xmath339 defined by @xmath340_+ e_j - e_k ) + x(\sum_{1\leq j \leq
m}[b_{kj}]_+ e_j - e_k ) , \\ \label{eq : mu2 } x'(e_j ) & = x(e_j ) \qquad 1\leq j
\leq m , j\neq k.\end{aligned}\ ] ] since @xmath341 , we see that @xmath342 is also unitally compatible .
it was constructed explicitly in ( * ? ? ?
* proposition 4.2 ) the automorphism @xmath328 for @xmath343 .
one should notice that the mutation @xmath30 is an involution .
let @xmath344 be the @xmath345-regular tree with root @xmath346 .
there is a unique way of associating a seed @xmath347 for each vertex @xmath348 such that 1 .
@xmath349 , 2 .
if @xmath350 and @xmath351 are linked by an edge @xmath13 , then the seed @xmath352 is obtained from @xmath353 by the mutation at @xmath13 . the quantum cluster algebra @xmath354 with initial seeds @xmath355 is the @xmath13-subalgebra of @xmath110 generated by all _ cluster variables _ @xmath356 , _ coefficients _ @xmath357 and @xmath358 .
recall the operators @xmath359 and @xmath360 . here
@xmath361 can also be written as the formal product @xmath362 so its inverse is the @xmath79-pochhammer symbol @xmath363 .
it satisfies @xmath364 from the fact that @xmath365 and @xmath366 we can easily deduce the following _ y - seeds _ mutation formula ( * ? ? ?
* ( 4.11 ) ) @xmath367_+e_k}\prod_{l=1}^{|b_{ik}|}(1+q^{\sgn(b_{ik})(l-\frac{1}{2})}y_k)^{\sgn(b_{ik } ) } & \text{otherwise}. \end{cases}\ ] ] we consider the lattice map @xmath368 .
this map induces a operator @xmath369 . by the unital compatibility of @xmath38 and @xmath45 , we have that @xmath370 .
so we conclude that the operator @xmath371 is an algebra homomorphism .
let @xmath372 be a sequence of edges connecting @xmath346 and @xmath373 .
we write @xmath374 for the sequence of mutation @xmath375 . for simplicity , we write @xmath376 for @xmath377 and @xmath378 for @xmath379 .
the next lemma says that the operator @xmath371 is compatible with mutations .
[ l : muxy ] @xmath380 for any @xmath381 . using the unitally compatibility of @xmath38 and @xmath45 ,
this is clearly reduced to prove for @xmath382 .
we prove by induction on @xmath62 .
for @xmath383 , it is trivial .
suppose that it is true for @xmath62 , then @xmath384_+e_k } \prod_{l=1}^{|b_{ik}^{s+1}| } ( 1+q^{\sgn(b_{ik}^{s+1})(l-\frac{1}{2})}y^{e_k})^{\sgn(b_{ik}^{s+1})}\big),\\ & = x_{s}(e_ib_{s}+[b_{ik}^{s}]_+e_kb_{s } ) \prod_{l=1}^{|b_{ki}^{s}| } \big(1+q^{\sgn(b_{ki}^{s})(l-\frac{1}{2})}x_{s}(e_kb_{s})\big)^{\sgn(b_{ki}^{s})}. \intertext{on the other hand , } x_{s+1}(e_ib_{s+1 } ) & = x_{s+1}(\sum_j b_{ij}^{s+1 } e_j),\\ & = \lambda_{s+1}(e_ib_{s+1}-b_{ik}^s e_k , b_{ik}^s e_k)x_{s+1}(\sum_{j\neq k } b_{ij}^{s+1 } e_j ) x_{s+1}(b_{ik}^{s+1 } e_k),\\ & = x_{s}(\sum_{j\neq k } b_{ij}^{s+1 } e_j ) \cdot \big(x_{s}(\sum_{1\leq
j \leq m } [ b_{jk}^{s}]_+e_j - e_k ) + x_{s}(\sum_{1\leq j \leq m } [ b_{kj}^{s}]_+ e_j - e_k)\big)^{b_{ki}^{s}},\\ & = x_{s}(\sum_{j\neq k } b_{ij}^{s+1 } e_j ) \cdot \big(x_{s}(\sum_{1\leq j \leq m } [ b_{jk}^{s}]_+e_j - e_k ) ( 1 + q^{\frac{1}{2}}x_{s}(e_kb_{s}))\big)^{b_{ki}^{s}},\\ & = x_{s}\big(\sum_{j\neq k } ( b_{ij}^{s}+[b_{ik}^{s}]_+b_{kj}^{s } - b_{ki}^{s}[b_{jk}^{s}]_+)e_j\big ) \cdot x_{s}\big(b_{ki}^{s}(\sum_{1\leq j \leq m } [ b_{jk}^{s}]_+e_j - e_k)\big)\cdot \\ & \qquad \prod_{l=1}^{|b_{ki}^{s}| } \big(1+q^{\sgn(b_{ki}^{s})(l-\frac{1}{2})}x_{s}(e_k b_s)\big)^{\sgn(b_{ki}^{s})},\\ & = x_{s}(e_ib_{s}+[b_{ik}^{s}]_+e_kb_{s } ) \prod_{l=1}^{|b_{ki}^{s}| } \big(1+q^{\sgn(b_{ki}^{s})(l-\frac{1}{2})}x_{s}(e_kb_{s})\big)^{\sgn(b_{ki}^s)}.\end{aligned}\ ] ] for any @xmath385 , there is a unique sequence of edges @xmath386 connecting @xmath346 and @xmath350 .
let @xmath0 be some _ non - degenerate _
( ( * ? ? ?
* definition 7.2 , proposition 7.3 ) ) potential of @xmath1 , and set @xmath387
. we shall assume the following condition for @xmath0 : for any @xmath385 , and any @xmath388 , the assumption @xmath34 holds for @xmath389 .
we do not know if such a potential exists for any @xmath44-matrix . to give another definition of @xmath390 for @xmath391
, we consider the extended qp @xmath52 from @xmath53 by adding a new vertex @xmath54 and @xmath55 new arrows from @xmath56 to @xmath54 .
we apply the inverse of @xmath48 to @xmath52 , and obtain a qp @xmath57 .
we freeze the same set of vertices of @xmath59 as that of @xmath1 , and let @xmath58 be the extended @xmath44-matrix of @xmath59 . although the extended vertex @xmath54 is not frozen , we will never perform mutation at @xmath54 henceforth . [
d : cluster ] for any @xmath391 , we define @xmath392 by theorem [ t : fwc ] , @xmath393 [ t : cluster ] definition [ d : cluster ] defines the quantum cluster algebra @xmath354 .
in particular , we have the following quantum cluster character @xmath394 we trivially extend @xmath38 to @xmath395 , and thus we have the natural embedding @xmath396 . note that @xmath397 , so they are _ not _ unitally compatible .
we say that they are unitally compatible on the principal part
. however , it still makes perfect sense if we define @xmath398 by and @xmath399 by the relation . clearly , for any @xmath400 , @xmath398 is also unitally compatible on the principal part .
this is all we need for the following analogue of lemma [ l : muxy ] holds : for each @xmath385 , we associate as before @xmath401 , then @xmath402 moreover , @xmath403 extends @xmath404 in the same way : @xmath405 so that we have the natural embedding @xmath406 for each @xmath348 .
hence , @xmath407 we can view @xmath408 as @xmath409 , where @xmath410 is the _ extended ( dual ) @xmath411-vector _ corresponding to the mutated cluster monomial .
[ ex : a3 ] consider quiver @xmath412 with potential @xmath413 .
we perform a sequence of mutations @xmath414 , and obtain the quiver @xmath415 with the same potential .
we choose @xmath416 as the cut .
it is easy to count the vanishing cycles for each dimension vector .
for example , for @xmath417 , @xmath418 note that @xmath419 counts neither the representation grassmannian of @xmath420 of the algebra + @xmath421 nor that of the algebra @xmath422 in particular , the condition ` numb ' can not be removed from lemma [ l : vtormu ] .
let @xmath62 be a sink of @xmath1 , and @xmath63 be a representation of @xmath1 .
we assume that @xmath63 does not contain the simple representation @xmath64 as a direct summand .
let @xmath423 we want to compare @xmath68 with @xmath424 .
we first consider the _ extension _ @xmath425:=\sm{kq & 0\\ m & k}$ ] of @xmath1 by @xmath63 .
this is an algebra of global dimension two , so we can complete it to a qp @xmath176 with a cut @xmath164 such that @xmath426 .
we freeze the extended vertex @xmath54 of @xmath174 , then @xmath427 . since all arrows in @xmath164 end in @xmath54 , by lemma [ l : vtormu ] , @xmath428}}\frac{|\rep_{(1,\beta)}^{\mu_\infty}(q[m])|}{|\gl_{(1,\beta)}|}x^{(1,\beta)},\ ] ] where @xmath429}$ ] is the euler form of @xmath430 $ ] , and @xmath25 is the framing stability .
@xmath431)$ ] can be identified with @xmath432 so the quotient @xmath431)/\gl_\beta$ ] is the representation grassmannian @xmath66 , and thus @xmath433)|}{|\gl_\beta||\gl_1|}x^{(1,\beta)},\\ & = \sum_{\beta } q^{-\frac{1}{2}\innerprod{\br{m}-\beta,\beta}}|\gr^\beta(m)| x^{(1,\beta)}.\end{aligned}\ ] ] the analogous statement does not hold for quivers with potentials in general ( see example [ ex : a3 ] ) . for any @xmath434 ( not necessary a sink ) , the cut @xmath164 of @xmath435 satisfies the conditions in corollary [ c : tilting ] , so we get another algebra @xmath436 .
the algebra @xmath437 is extended from @xmath438 by @xmath439 .
let @xmath54 be the extended vertex of @xmath440 $ ] , and @xmath441 .
then @xmath442 since there is no incoming arrows to @xmath54 for @xmath440 $ ] , we have that @xmath443 .
let @xmath444 be the bb - tilting module of @xmath174 at @xmath62 , and @xmath445 .
we need to show that @xmath446 let @xmath447 be the bb - tilting module of @xmath85 at @xmath62 . since @xmath63 does not contain @xmath64 as a direct summand the quiver of @xmath440 $ ] has no arrow from @xmath54 to @xmath62 .
so @xmath448 , and thus @xmath449 .
hence , @xmath450 .
for the second one , we consider @xmath451 [ t : qgrass ] @xmath68 and @xmath452 are related via .
in particular , if @xmath63 is _ polynomial - count _ , that is , all its grassmannians @xmath66 are polynomial - count , then so are all reflection equivalent classes of @xmath63 .
consider the following quiver @xmath453^{b_1,b_2,b_3 } \\ 1 \ar[ur]^{a_1,a_2,a_3 } & & 3 \ar[ll]^{c_1,c_2,c_3 } } } \ ] ] with potential @xmath454 and cut @xmath455 .
then the algebra @xmath163 is extended from the quiver @xmath456 by a representation @xmath63 of dimension @xmath457 ( see ( * ? ? ?
* example 7.8 ) ) . to compute @xmath68 , we can first compute @xmath458 , where @xmath459 can be presented by the following base diagram @xmath460 the black ( resp .
white ) dots are a basis at vertex 3 ( resp .
vertex 2 ) ; the letter on an arrow represents the identity map on the arrow of the same letter .
@xmath461(x^{(1,0,1)}+x^{(1,0,2)}+x^{(1,1,3)}+x^{(1,2,3)})+3q^{\frac{1}{2}}x^{(1,1,2)}.\ ] ] here @xmath462 $ ] is the quantum number @xmath463 . using theorem [ t : qgrass ]
, we find that @xmath464(x^{(1,1,0)}+x^{(1,2,0)}+x^{(1,2,3)})+3q^{\frac{1}{2}}x^{(1,1,1)}\\ & + [ 3][3](x^{(1,2,1)}+x^{(1,2,2)})+(q^{\frac{3}{2}}+q^{-\frac{3}{2}})[3](x^{(1,3,1)}+x^{(1,3,5)})\\ & + ( q-1+q^{-1})[3][5](x^{(1,3,2)}+x^{(1,3,4)})+(q-1+q^{-1})[4][5]x^{(1,3,3)}.\end{aligned}\ ] ] employing the methods developed in @xcite , we can compute @xmath465 and @xmath466 for all @xmath28 with @xmath467 and generic @xmath29 .
theorem [ t : tilting ] generalizes the main result of @xcite from apr - tilting modules to bb - tilting modules .
we slightly simplify its proof as well .
* proposition 3.3 ) [ l : presentation ] let @xmath1 be a finite quiver and @xmath174 be a finite dimensional basic algebra .
let @xmath468 be a set of relations , then the following are equivalent 1 .
@xmath174 can be presented as @xmath469 .
2 . there is an algebra homomorphism @xmath470 such that the sequence @xmath471 is exact .
we set @xmath472 and @xmath473 .
recall that the bb - tilting module @xmath251 can be presented as @xmath474 where @xmath254 and @xmath475 . using the presentation @xmath169 of @xmath163 .
we have for @xmath270 @xmath476 where @xmath477 and @xmath478 .
since the cut @xmath164 satisfies definition [ d : exact].(2 ) , the first three terms are part of the minimal projective resolution of @xmath74 .
we assume that the projective @xmath479 is minimal as well .
this fits in the following commutative diagram @xmath480 & \displaystyle c_{ki } p_k \ar[r]^{\oplus \alpha } \ar@{^{(}->}[d]^{\iota } & \displaystyle c_{ki } p_{in } \ar[r]^{\oplus g } \ar[d]^{\partial_{acb } } & \displaystyle c_{ki } t_k \ar[r ] \ar[d]^{f:=_c(f_c ) } & 0 \\
p'\ar[r]^{h } & \displaystyle
\bigoplus_{hc = i , c\in c } p_{tc } \ar[r]^{\partial_{cb } } & \displaystyle \bigoplus_{tb = i } p_{hb } \ar[r]^{\beta } & p_i \ar[r ] & s_i}\ ] ] here @xmath481 , and the first row is a direct sum of @xmath482 copies of ; the map @xmath483 is the natural embedding , the map @xmath484 is given by the matrix @xmath485 , and @xmath4 is induced from @xmath484 .
we then take the mapping cone of the above diagram , and cancel out the last term @xmath486 .
we end up with @xmath487 let @xmath488 be the quiver obtained from @xmath489 by forgetting all arrows in @xmath258 . to apply lemma [ l : presentation ] , we construct an algebra homomorphism @xmath490 as follows . for any direct summands @xmath491 of @xmath249 , we will view @xmath492 under the natural embedding into @xmath493 .
let @xmath494 be the identity map in @xmath495 .
we define 1 .
@xmath496 , 2 .
@xmath497 for @xmath498 , 3 .
@xmath499 , 4 .
@xmath500 for @xmath262 , 5 . @xmath501)= ab\in\hom_a(p_{ha},p_{tb})$ ] for @xmath261 .
recall that @xmath502+\sum_{ha = tb = k}b^*a^*[ab],$ ] and @xmath258 contains all 1 .
@xmath259 if @xmath260 , 2 .
arrows @xmath245 if @xmath261 , 3 . composite arrows
@xmath240 $ ] with @xmath262 .
so the corresponding relations @xmath503 are given by 1 .
@xmath504\}_{tc\neq k}$ ] , 2 .
@xmath505\}_{b\notin c}$ ] , 3 .
@xmath506 .
hence , theorem [ t : tilting ] is the consequence of the following two lemmas .
[ l : res1 ] we have the following exact sequence @xmath507 where @xmath508 is the matrix @xmath509 .
we apply @xmath272 to the exact sequence , and get @xmath510 the last term @xmath511 vanishes because the first map below is surjective @xmath512 next , @xmath513 is one - dimensional because of the following exact sequence @xmath514 finally , we claim the image of @xmath515 is exactly the image of @xmath516 . by the definition of @xmath508
, it suffices to show that @xmath517 is surjective .
but the cokernel of @xmath518 is @xmath519 . applying @xmath272 to ,
we get the complex @xmath520 [ l : res2 ] this complex is exact and induces @xmath521 we first show that the complex is exact at @xmath522 .
we apply @xmath272 to the exact sequence @xmath523 and get @xmath524 if @xmath525 such that @xmath526 , then @xmath527 . so it suffices to show that @xmath528 .
the condition @xmath529 implies that @xmath479 has no @xmath530 as its summands .
so @xmath531 , and thus @xmath532 since @xmath533 has no @xmath530 as its direct summands , for the same reason the complex is exact at @xmath534 .
we remain to show that the cokernel of @xmath535 is one - dimensional .
let @xmath536 be the first syzygy of @xmath74 .
we apply @xmath272 to @xmath537 and obtain @xmath538 since @xmath539 vanishes , the cokernel of @xmath535 is the same as that of @xmath540 . by applying @xmath541 to
, we see that @xmath542 in the meanwhile , @xmath543 so @xmath544 together with , we conclude that the cokernel of @xmath535 is @xmath13 . finally , we prove corollary [ c : tilting ] . since the cut satisfies , there is no relation starting from @xmath545 .
so @xmath545 has projective dimension one , and we have that @xmath546 .
hence @xmath547 , and the map @xmath515 in lemma [ l : res1 ] is in fact injective .
now @xmath163 has global dimension 2 , so @xmath479 in is zero , and thus the map @xmath548 of lemma [ l : res2 ] is injective .
we conclude that @xmath271 has global dimension 2 as well .
the two resolutions of lemma [ l : res1 ] and [ l : res2 ] also imply that @xmath549 is a minimal set of generators in @xmath550 .
the author thanks mathematical science research institute in berkeley ( msri ) for its hospitality and support during the research program cluster algebras of fall 2012 when most of results are obtained
. he also wants to thank fan qin for patiently answering several questions , and professor bernhard keller for his encouragement .
99 c. amiot , s. oppermann , _ cluster equivalence and graded derived equivalence , _ arxiv:1003.4916 .
i. assem , d. simson , a. skowroski , _ elements of the representation theory of associative algebras , _ london mathematical society student texts 65 , cambridge university press , 2006 .
k. behrend , j. bryan , b. szendri , _ motivic degree zero donaldson - thomas invariants , _ invent .
192 ( 2013 ) , no .
1 , 111160 .
a. b. buan , o. iyama , i. reiten , d. smith , _ mutation of cluster - tilting objects and potentials , _ amer . j. math . 133 ( 2011 ) , no .
4 , 835887 .
a. berenstein , a. zelevinsky , _ quantum cluster algebras , _ adv .
( 2005 ) , no .
2 , 405455 . b. davison , d. maulik , j. schuermann , b. szendri , _ purity for graded potentials and cluster positivity , _ preprint
. h. derksen , j. weyman , a. zelevinsky , _ quivers with potentials and their representations i. mutations , _ selecta math .
14 ( 2008 ) , no .
1 , 59119 . h. derksen , j. weyman , a. zelevinsky , _ quivers with potentials and their representations ii .
applications to cluster algebras , _ j. amer .
( 2010 ) , no .
3 , 749790 .
a. i. efimov , _ quantum cluster variables via vanishing cycles , _ arxiv:1112.3601 .
j. fei , _ counting using hall algebras ii .
extensions from quivers , _ arxiv:1302.1835 .
t. hausel , f. rodriguez - villegas _ mixed hodge polynomials of character varieties , _ invent . math . 174 ( 2008 ) , no .
3 , 555624 . m. herschend , o. iyama _ selfinjective quivers with potential and 2-representation - finite algebras , _ compos . math . 147 ( 2011 ) , no
. 6 , 18851920 .
b. keller _ deformed calabi - yau completions , _ j. reine angew . math . 654 ( 2011 ) , 125180 . b. keller , _ on cluster theory and quantum dilogarithm identities , _ representations of algebras and related topics , 85116 , ems ser .
, eur . math .
zrich , 2011 .
king , _ moduli of representations of finite - dimensional algebras , _ quart .
oxford ser .
( 2 ) 45 ( 1994 ) , no .
180 , 515530 .
m. kontsevich , y. soibelman , _ cohomological hall algebra , exponential hodge structures and motivic donaldson - thomas invariants , _
s. ladkani , _ perverse equivalences , bb - tilting , mutations and applications , _ arxiv:1001.4765
. y. mizuno , _
apr tilting modules and graded quivers with potential , _
s. mozgovoy , _ on the motivic donaldson - thomas invariants of quivers with potentials , _ arxiv:1103.2902 .
k. nagao , _ wall - crossing of the motivic donaldson - thomas invariants , _
m. reineke , _ the harder - narasimhan system in quantum groups and cohomology of quiver moduli , _ invent .
( 2003 ) , no .
2 , 349368 .
o. schiffmann , _ lectures on hall algebras , _ arxiv:0611617 . | for a quiver with potential , we can associate a vanishing cycle to each representation space .
if there is a nice torus action on the potential , the vanishing cycles can be expressed in terms of truncated jacobian algebras .
we study how these vanishing cycles change under the mutation of derksen - weyman - zelevinsky .
the wall - crossing formula leads to a categorification of quantum cluster algebras under some assumption .
this is a special case of a. efimov s result , but our approach is more concrete and down - to - earth .
we also obtain a counting formula relating the representation grassmannians under sink - source reflections .
[ multiblock footnote omitted ] |
the search to directly image an extrasolar planet requires contrast levels of @xmath010@xmath1 a few @xmath2 from the central star .
scattered light in a telescope and the diffraction pattern of the telescope s aperture limit the contrast possible for direct detection of faint companions .
the circular aperture of telescopes creates a sub - optimal diffraction pattern , the so - called airy pattern which is azimuthally symmetric .
in addition , the intensity in the diffraction pattern of the circular aperture declines as @xmath3 , where @xmath4 . currently the best way to diminish the airy pattern is to use a coronagraph by using the combination of a stop in the focal plane that rejects a majority of the central bright object s light and a lyot stop in the pupil plane to reject high frequency light ( lyot 1939 ; malbet 1996 ; sivaramakrishnan et al . 2001 ) .
several recent ideas explore the use of alternative `` apodized '' apertures for high contrast imaging in the optical or near - infrared ( nisenson & papaliolios 2001 ; spergel 2002 ; debes , ge , & chakraborty 2002 ) . these designs revisit concepts first experimented with in the field of optics ( jacquinot & roizen - dossier 1964 ) .
other designs , such as the band limited mask , seek to null the light from a central star in much the same way that a nulling interfermoeter performs ( kuchner & traub 2002 ) . by placing a mask into the pupil plane with a gaussian aperture
, one can transform a traditional circular aperture telescope into one with a diffraction pattern better suited for high contrast imaging . using
a mask represents a quick , efficient , and cheap way to test this emerging imaging method to determine its advantages and tradeoffs and compare them to the performance of other existing techniques .
preliminary results of observign with a prototype gaussian pupil mask can be found in debes et al .
( 2002 ) . in this
proceeding we report further lab tests of the performance of gapms compared to lyot coronagraphs with similar throughput as well as testing the new technique of combining a gapm with a coronagraphic image plane mask .
we theoretically compare different techniques with the same throughput on the hubble space telescope to determine what may be useful in a real spacecraft .
a fairer comparison between a lyot coronagraph and the gapm is to have equal throughput designs and compare their contrast levels . as part of lab experiments that we are performing
we compared two new gapm designs with a lyot coronograph that had a comparable throughput .
two types of designs were tested , idealized apertures with no secondary structure ( 20% throughput ) and realistic masks that will be used for future observing ( 30% throughput ) , which have two gaussian aperture per quadrant , avoiding support structure .
unsaturated images gapms were taken to predict the flux for the longer , saturated images .
we took exposures that had on the order of 10@xmath5 counts in order to image the psf .
we found that this was insufficient to get high s / n on the fainter portions of the psf and we estimate that beyond 1 - 2@xmath6 the read noise begins to dominate .
longer integrations are planned . for the coronagraphic modes we used a gaussian transmission focal plane mask with a fwhm of 500@xmath7 ( @xmath8 ) .
short exposures were taken without the mask for estimating the peak flux of an unblocked point source for a given exposure time .
the mask was then carefully aligned to within 1 pixel to block the point source .
figure 1 shows the results of these lab tests . in both cases
the hybrid designs perform as well or better , both reaching @xmath9 .
the flattening of the profile suggests that the observations were hitting the read noise limit of the picnic detector on piris .
other experiments that were done where the mask was misaligned by several pixels presented dramatically worse results , underscoring the need for subpixel alignment and stability over an observation .
this points to the utility of the gapms alone for quick surveys and hybrid or lyot designs for deeper searches .
given the large number of potential ideas for high contrast imaging , and due to a lack of many lab tests of these designs , a way to model some of the various contrast degradations present in a real space mission would be useful for determining what designs are better suited for future tpf type missions .
a strong test would be to compare the performance of the different designs on the hst , where many of these different errors are well modeled .
tinytim , the psf modeling software used by the space telescope institute , has accurate wavefront error maps of the telescope ( krist 1995 ) . using these wavefront error maps as input to our models allows us to compare all of the different designs with an equal footing of throughput .
figure 2 shows a horizontal cut along the axis of highest contrast for all of the designs . in this preliminary simulation
the band limited mask performs slightly better than the other designs for the same amount of throughput .
this is because the lyot stop for the band limited mask is undersized from its optimal shape by about 10@xmath10 .
this would block some of the residual light that leaks through the focal plane mask .
these simulations ignore mask errors in both the focal and pupil planes .
possibly with psf subtraction techniques a factor of 10 - 100 deeper contrast could be achieved , allowing some bright extrasolar planets to be observed with hst .
further simulations need to be done to better understand the feasibility of such observations .
we have performed several simulations , lab tests , and telescope observations with gapms and lyot coronagraphs in order to better understand the interplay between theory and the reality of observations .
gapms alone provide an improvement over a simple circular aperture for quick high contrast imaging .
the combination of a gapm and a coronagraphic mask further supresses and improves the performance of the gapm alone .
the masks are very sensitive to an accurate reproduction of shape and thus need accuracies that may be as restrictive as sub - micron precision .
this is possible with new nanofabrication techniques that have been perfected at the penn state nanofabrication facility , where future masks may be produced . precisely fabricating these masks
can potentially improve performance to the ideal limit for a mask provided it is above the scattered light limit of the telescope , bringing it in line with lyot coronagraphs of comparable throughput .
simulations of different techniques on the hst provide an avenue to test new technologies for future nasa missions such as tpf , and show what can be possible from space .
j.d acknowledges funding by a nasa gsrp fellowship under grant ngt5 - 119 .
this work was supported by nasa with grants nag5 - 10617 , nag5 - 11427 as well as the penn state eberly college of science . j.
g. also acknowledges funding through ball aerospace co .. | gaussian aperture pupil masks ( gapms ) can in theory achieve the contrast requisite for directly imaging an extrasolar planet .
we use lab tests and simulations to further study their possible place as a high contrast imaging technique .
we present lab comparisons with traditional lyot coronagraphs and simulations of gapms and other high contrast imaging techniques on hst .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
with the decade - long work of sensitive x - ray observatories and , a refined picture of stellar x - ray emission in our galaxy is now available ( for a review , see e.g. , * ? ? ?
however , many x - ray production processes in stars depend on metallicity , and this dependence has not yet been tested thoroughly . in this context , the magellanic clouds provide an opportunity to observe the metallicity effects . for example , the tentative discovery of x - ray emission from low - mass pre - main sequence ( pms ) objects in the small magellanic cloud ( smc ) was reported recently and the emission level appears comparable to that in the galaxy , constraining emission models @xcite . on the high - mass end ,
a 26ks exposure of the emblematic region 30 doradus was reported by @xcite , leading to the detection of 180 x - ray sources - a hundred being found in the massive central cluster r136 , with some of these sources displaying @xmath52 net counts and thus having a non - negligible probability to be spurious . from their spectral analyses , @xcite derived for the massive stars of 30 dor absorption columns of 110@xmath6h - atom@xmath7 , temperatures of 0.54kev , and absorption - corrected x - ray luminosities from @xmath8ergs@xmath1 down to @xmath9ergs@xmath1 ( the sensitivity limit ) .
@xcite suggested that many , but not all , of the detected massive stars were colliding - wind binaries , which could be ascertained through further monitoring .
furthermore , these authors found no clear ratio for their sample , contrary to that for the galaxy ( @xmath10 , see e.g. * ? ? ? * and references therein ) .
this conclusion may be reconsidered , however , because ( 1 ) a significant contamination by the x - ray bright colliding wind binaries ( esp .
wr+o ) may hide trends intrinsic to individual massive stars ( but correcting this problem needs both optical and x - ray extensive monitoring , which is not available ) and ( 2 ) coherent ratios are found only when x - ray luminosities are well measured statistically and are corrected by the _
interstellar _ absorption , not by the _ total _ absorption .
it should also be underlined that 30 dor is an extreme environment , more akin to starbursts than a good representative of the large magellanic cloud ( lmc ) population . in the lmc , n11 is the second largest region , just after the giant region 30 dor .
the less extreme properties of n11 make it much more representative of lmc regions and clusters of massive stars . besides , the lower concentration of stars implies less source confusion ,
hence should lead to more reliable results .
star formation has been very active in n11 , with no less than four recognized ob associations : lh9 , lh10 , lh13 , and lh14 @xcite .
the stellar feedback has restructured the surrounding interstellar material .
notably , the winds and supernova explosions of the massive stars in lh9 have gradually carved a cavity , giving rise to an expanding superbubble some 120 pc in size @xcite .
this expansion has probably triggered the formation of lh10 at the periphery of the superbubble @xcite . in turn
, the massive stars of lh10 are now beginning to blow bubbles @xcite , further triggering new star formation in their surroundings @xcite .
n11 is clearly one of the best sites to study the interplay between stars and the interstellar medium .
this interaction , often violent , produces x - ray emission . using a 30 ks _
position sensitive proportional counter observation , @xcite reported the first detection of x - rays in n11 .
this _ rosat _
observation revealed the presence of extended areas of diffuse emission , with the brightest sources associated with the n11l supernova remnant ( snr ) and within the superbubble around lh9 .
further investigation was performed with a 30 ks _
xmm - newton _ observation , which provided the first detection of point sources in the field @xcite .
while stars in lh9 remained unresolved , this _ xmm _ observation unveiled in lh10 a mixture of diffuse emission and point sources associated with some of the most massive stars of the cluster .
a detailed x - ray analysis of n11 requires a combination of both high sensitivity and high spatial resolution , which became possible with our new , deep x - ray investigation of n11 , using the _ chandra x - ray observatory_. this observation will lead to several analyses , and this first paper discusses the point source population .
the aim is to uncover the nature of the point sources , and to find whether the properties of the stars in n11 differ from similar objects in the galaxy .
this paper is organized as follows : section 2 presents the data and their reduction , section 3 introduces the catalog of x - ray sources and their global properties , section 4 discusses extended and point sources associated with massive stars .
finally , section 5 summarizes our results .
the _ chandra _ acis - i observations of n11 were made in six separate segments within two months in 2007 . as summarized in table[obs ] , the exposure time of each segment was 4249ks and the roll angle ranged from 130@xmath11 to 188@xmath11 .
the pipeline products of the observations were reduced and analyzed using our own idl - based analysis tools ( e.g. , @xcite ) as well as the official software for the interactive analysis of observations ( ciao ; version beta1 - 4.0 ) together with the calibration database ( caldb 3.4.0 ) and other publicly available routines ( e.g. , xspec v 12.7.0 ) . it should be noted that our detection procedure was shown to give results comparable to those of acisextract @xcite , and this was checked on this dataset by a quick run of that tool .
we used the light - curve cleaning routine `` lc@xmath12clean '' to remove time intervals of significant background flares when count rates deviated more than 3@xmath13 or a factor of @xmath14 from the mean rate of individual observations .
this cleaning , together with a correction for the dead time of the six observations , resulted in a total of 280ks useful exposure for subsequent analysis .
a combination of source detection algorithms ( wavelet , sliding - box , and maximum likelihood centroid fitting ) were applied to unsmoothed data in three bands : soft ( s ) 0.52kev , hard ( h ) 28kev , and total ( t ) 0.58kev @xcite . briefly
, the wavelet detection was first used to find the initial source candidates with a high threshold of local false detection probability @xmath15
. then , background maps were constructed by removing the wavelet - detected sources and by conducting a median averaging or smoothing of the three input images on scales much greater than the point spread function ( psf ) to achieve an intensity uncertainty @xmath16 .
these background maps , insensitive to the exact details of the construction procedure , were used to search again for sources , using this time a sliding - box algorithm ( the so - called map detection mode ) .
finally , a maximum likelihood centroiding algorithm was used , still using the count background map , to derive the best centroid positions for the sources @xcite .
this process was applied independently to each energy band .
our final source list contains sources with local false detection probability @xmath17 at least in one band ( poisson statistics was used in calculating the significance of a source detection above the local count background ) .
the sensitivity of the source detection depends on the size of the psf as well as the local background level and effective exposure , which all vary with position , especially with the off - axis angle of the detected sources .
the source detection , though optimized for point - like sources , includes a few strong peaks of the diffuse x - ray emission , chiefly associated with the snr n11l , about @xmath18 west of the field center ( for more on this object , see sun et al . in preparation ) .
some of these sources associated with peaks of the diffuse emission are detected in s - band only .
once source positions were identified , source count rates needed to be estimated . to this aim , it must be recalled that the most precise effective exposure times are evaluated in narrow energy bands .
therefore , we calculated the net ( background - subtracted ) count rates in four subbands ( s1=0.51kev , s2=12kev , h1=24kev , h2=48kev ) , and they were later added to form the rates in the broader bands ( s , h , and t ) .
we thus first constructed effective exposure maps in these four subbands .
the construction of these exposure maps assumed a power law spectrum of photon index 1.7 and accounted for the telescope vignetting and bad pixels as well as the quantum efficiency variation of the instrument , including the time - dependent sensitivity degradation , which is particularly important at low energies ( @xmath19kev ) .
[ fig : exp_1 ] shows such a merged effective exposure map , illustrating the features of the bad pixel removal , ccd gaps , observation dithering , etc . as well as the overall field coverage .
in order to treat uniformly both strong and weak sources , source counts for each subband were then extracted within the 70% energy - encircled radius ( eer ) of the psf , whose size depends on the off - axis angle of the source in the exposure and of the energy band under consideration . a background correction using the background map constructed earlier
was applied .
finally , count rates were derived from dividing source net counts by their effective exposure times ( values at the source positions in the exposure map of the energy band under consideration ) , leading to equivalent on - axis values .
it should be noted that the presented count rates have thus been corrected for the full psf and for the effective exposure , which accounts for not only the telescope vignetting , but also the degradation of the detector sensitivity with time .
therefore , the actual number of counts in a detection aperture is not simply a count rate multiplied by an exposure of 280ks
. the difference could be up to a factor of @xmath20 , depending on a source s spectral shape .
we extract an acis spectrum for each source detected with @xmath21 .
the on - source spectral extraction circle has the same radius as used for the source removal , while the local background spectrum is estimated from a concentric annulus with the inner radius equal to 2@xmath22 the circle radius and the outer radius twice larger .
detected sources are removed from the background region .
the background spectrum is normalized accounting for bad pixels and boundaries of the ccds as well as the source removal .
we obtain the averaged response matrices of each source spectrum , using the weights derived from on - source 0.52kev band counts in the detector coordinates of individual observations .
the spectrum is further adaptively binned to achieve a background - subtracted signal - to - noise ratio greater than 2.5 in each bin .
we compared the positions of a few well identified x - ray sources with their ( known ) optical counterparts . to this aim , we considered only ob stars , since they are rather bright sources of x - rays , whereas no other stellar x - ray emitter was a priori known in the field ( though several other sources may have possible stellar counterparts , see below , but these were not a priori known ) .
we found no significant systematic offset ( @xmath23 05 ) and therefore , no astrometry correction was applied to the x - ray data .
we caution however that these sources lie at large off - axis angles , so that the uncertainties in their x - ray positions may be large , but the absence of bright x - ray sources with well - established optical counterparts prohibits us to fine - tune the astronometry to 01 accuracy .
with the detection procedure described in the previous section , we found 165 sources in n11 : 43 of them were detected with the highest confidence or smallest @xmath24 value in the s band , 5 in h band , and 117 in t band . amongst these ,
74 were detected in all three bands , 56 only in two bands , and 35 only in one band ( 22 for s , 2 for h , 11 for t - while the total band often maximizes the signal - to - noise ratio , some very soft or very hard sources are more easily detected in only the soft or hard band , for example , nearby stars and diffuse emission peaks in the soft band , and faraway accreting sources and agns in the hard band ) . table [ sourcelist ] lists for each detected source its position , count rate in the total band and hardness ratios @xmath25 , and @xmath26 , as well as its off - axis angle , number of counts , estimated background counts and _ effective _ exposure in the detection aperture and in the total band .
the last column yields the band where the source was detected with the highest confidence .
while count rates may look large at first sight , it must be recalled that the source detection completeness varies across the acis - i map , depending on the local psf , effective exposure ( fig .
[ fig : exp_1 ] ) , and background .
furthermore , count rates have been corrected for ccd degradation , leading even more to apparently high count rates . without making this correction ,
the interpretation of the hardness ratios or the count rate to flux conversion given below would depend on the time when observations were taken . the lowest value for source net counts and signal - to - noise ratio are 4.7 and 2.1 , respectively .
the conversion from a count rate to an unabsorbed energy flux depends on the source spectrum and foreground absorption .
a characteristic value of the conversion is @xmath27 @xmath28 in the 0.58kev band for a power - law spectrum of photon index 2 and an absorbing - gas column density @xmath29h - atom@xmath7 ( assuming solar abundances ) .
this conversion should be a good approximation ( within a factor of 2 ) for @xmath30h - atom@xmath7 .
the corresponding conversion to a source - frame luminosity in the same band is @xmath31 @xmath32 at the lmc distance of 50 kpc .
the acis - i total band ( unsmoothed ) image is shown in fig .
[ fig : x14_sou ] with the detected x - ray sources marked , while fig.[fig : rgb ] shows a smoothed three - color map of the x - ray emission in n11 as well as an h@xmath33 image for comparison .
the smoothed images show that many point sources are superposed on diffuse emission . as the source detection
is based on poisson statistics and the image smoothing uses gaussian statistics , there appear to be additional point - like sources ( to eyes ) in the smoothed images .
these are artifacts of the smoothing procedure .
we will not consider these false sources caused by noise bumps .
we also conducted tests for timing variability .
we first carried out kolmogorov - smirnov tests as well as @xmath34 tests on the total band light curves of the 41 sources with detected @xmath35 , which are well covered by the six observations . for @xmath34 tests
, lightcurves were adaptively binned so that each bin contains at least 20 counts .
sources [email protected] and [email protected] ( # 32 and 158 , fig .
[ fig : timing ] ) show significant variabilities in the total band at confidence levels of 5 and 3@xmath13 , respectively .
they possess no counterpart within 1 ( see sect 3.1 ) , and their nature remains unknown .
variability examination in the s and h bands yields no additional results .
the analysis of individual observations separately , for all 117 sources that are not near the ccd gaps ( with a 12 margin to account for the dithering effect ) in any of the observations , yields only one positive result : [email protected] ( # 64 ) shows an apparent variation in the s band during the observation # 8210 and at a confidence level of @xmath37 ( fig .
[ fig : timing ] ) .
this source is a known quasar candidate ( see sect . 3 ) .
however , with so many sources studied in two independent bands ( the third one being related to the other two since @xmath38 ) , this latter variability detection is not inconsistent with the occurrence of such an event by pure chance hence is marginal .
figure [ hr ] shows hardness ratios of the 53 sources with both ratios known with errors less than 0.2 : most ( 48 ) sources have @xmath39 and @xmath40 , indicating relatively hard sources . to assess the contamination of the catalog by background agns , we have characterized the x - ray source number - flux relation ( nfr ) in n11 .
this nfr analysis uses only the 141 sources detected in the total band , for homogeneity and to avoid biases towards soft or hard sources .
eddington bias was corrected following the approach of @xcite .
we may compare the derived nfr to the log(@xmath41)log(@xmath42 ) presented by ( * ? ? ? * and references therein ) .
however , the x - ray absorption through n11 ( from nrao survey , @xmath43h - atom@xmath7 ) is substantially higher than that toward the deep fields ( foreground absorption @xmath44 ) . correcting for this difference
, we find an expected number of agns in the field of 91 ( fig .
[ nfr ] ) , i.e. , most of our sources are in fact extragalactic background objects .
indeed , a few agns have been identified in the field : a correlation of our source list with quasar tables in vizier results in the identification of sources # 11 @xcite and # 38 , 45 , 64 , 98 , 117 , 123 @xcite .
note however that some parts of the nebula are filled with molecular clouds , and these locally enhanced absorption columns may reduce the number of detectable agns .
counterparts to our x - ray sources were searched in several catalogs : usno - b1.0 catalog @xcite , guide star catalog v2.3.2 ( gsc , * ? ? ?
* ) , 2mass all - sky catalog of point sources @xcite , magellanic clouds photometric survey ( mcps , * ? ? ?
* ) , irsf magellanic clouds point source catalog @xcite , denis catalogue toward magellanic clouds ( dcmc , * ? ? ?
* ) , and @xmath45 photometry of n11 young stellar objects ( [ hkn2006 ] , @xcite ) . to find the optimal correlation radius ,
we first searched for the closest counterpart to each x - ray sources and derived the number of matches as a function of radius : at large radii , the number of matches is proportional to the squared radius , as expected by chance coincidences . a best correlation radius of 1 was found , and used to derive the final list of counterparts ( table [ counterparts ] ) : 71 of the 165 sources have at least one counterpart within 1 . amongst these
, thirteen objects are known massive stars ( see table [ deto ] and sect .
4 ) , two are ob candidates ( # 130 and 136 , @xcite ) , and one is a haebe candidate ( # 63 , @xcite ) .
two additional sources have been misidentified with stars in the past : src # 98 , proposed to be a haebe candidate on the basis of the photometry @xcite , is in fact a quasar @xcite , while src # 24 , identified to be a young stellar object @xcite , actually corresponds to the nucleus of a background galaxy @xcite .
the photometric measurements of the counterparts appear coherent in different catalogs .
we therefore focus on irsf , because it contains the largest number of counterparts ( 58 sources ) amongst the tested catalogs . considering the sources with full @xmath45 photometry available , color - magnitude and color - color diagrams
can be constructed ( fig .
[ hr ] ) . besides massive stars , counterparts appear to the right of the main sequence , suggesting that they are young , still forming stars
; however , known quasars also have similar photometric properties , requiring additional investigation . to this aim ,
we further used h@xmath33 , [ ] , and [ ] images taken with the mosaic camera on the blanco 4 m telescope at the cerro tololo inter - american observatory . the h@xmath33 observations consisted of three dithered exposures of 300s each for each location ; the bulk of n11 was imaged on 2008 december 5 and the periphery of n11 was imaged on 2010 january 9 .
the [ ] @xmath465007 observations and the [ ] @xmath476716,6731 observations consist of four dithered 450s exposures for each location ; these images were obtained on 2011 october 31 .
we have also used infrared observations made with the _ spitzer space telescope_. the _ spitzer _ images and photometry of point sources of n11 are taken from the previous work by @xcite , who made a photometric catalog and identified young stellar objects for the entire lmc .
these images have been used for inspecting the counterparts to the x - ray point sources . among the three optical images , the [ ] image is the most useful because the diffuse emission from ionized gas is not as strong and confusing as that in h@xmath33 and [ ] . among the infrared images ,
we have used primarily the 3.6 and 8.0 @xmath48 m images .
figure [ map ] shows @xmath49 cutout [ ] images overplotted with error circles ( radius from column 3 of table 2 ) centered on the derived positions of all 165 x - ray sources .
these images are useful for an independent confirmation or rejection of optical counterparts .
although some of the error circles have radii @xmath50 , the combined errors in the x - ray and optical astrometry may reach @xmath51 .
we have conservatively use a minimum correlation radius of 1@xmath52 for identification of optical counterparts ; however , only 3 sources ( # 43 , 78 , and 107 ) have optical counterparts that are within 1@xmath52 but outside the error radius . the results of our investigation are noted in table [ addcount ] , including _ spitzer _ counterparts of the x - ray point sources .
to assess the physical nature of the optical and infrared counterparts of the x - ray sources , we have assembled spectral energy distributions ( seds ) for the sources that have photometric measurements available .
the following passbands and catalogs have been used : @xmath53 from mcps , @xmath45 from 2mass and irsf , and irac bands from @xcite .
these seds are presented in figure [ sed ] .
the seds of known massive stars , such as sources # 48 , 77 , 97 , 110 , 115 , 126 , 141 , 142 , 146 , and 157 , have a distinct shape that falls off toward long wavelengths , following the rayleigh - jean law .
while the seds of stars can be diagnosed by their downturn in infrared , the nature of the stars need to be estimated from their photometric colors and magnitudes . for each object with a stellar sed ,
we use the mcps @xmath53 and irsf @xmath54 photometry to determine colors in several combinations of bands ( such as @xmath55 , @xmath56 , @xmath57 , @xmath58 , etc . ) , compare the observed colors with those of dwarfs ( luminosity class v ) and supergiants ( luminosity class i ) to assess its spectral type , and compare the observed magnitude with the expected absolute magnitude to determine its distance .
we find that sources # 130 and 136 are b2 giants , in agreement with the suggestion of candidate ob stars by hatano et al .
( 2006 ) , sources # 108 , 139 , and 145 are late - type b dwarfs , and source # 106 may be of type a0 .
distances that we derived for these objects are @xmath5950 kpc , and they are thus in the lmc . using the same conversion factors as for known massive stars ( see sect .
4.2 ) , the x - ray luminosities of these sources , in the total band , amount to @xmath60ergs@xmath1 for the three late - b and the a0 star , and @xmath61ergs@xmath1 for the two b2iii stars .
this leads to of @xmath62 for the a0 star , @xmath63 for the late - b stars , and @xmath64 for the b2iii stars .
such luminosities are too high for even flares of pms companions or flares of these stars themselves @xcite , but too low in comparison with those of known hmxbs in the lmc ( @xmath65ergs@xmath1 , * ? ? ?
furthermore , such high ratios are not compatible with embedded wind shocks ( such stars would not have the strong mass - loss necessary for this mechanism , anyway ) . in the absence of further information
, it seems likely that objects other than these ab stars are the true x - ray emitters : chance alignment with foreground ( soft ) source or background ( hard ) source or localized peak in soft diffuse emission . in this context , it should be noted that # 106 , 108 and 130 seem to emit mostly soft x - rays , while the emissions of # 136 and 139 appear much harder .
the cooler stars are all in the galactic foreground : source # 56 is a known k0 star ; sources # 5 and 28 are g stars ; 25 , 43 , 78 , 87 , and 109 are k stars ; 54 , 95 , and 107 are m stars . in two cases ,
sources # 25 and 109 , the colors at long wavelengths suggest k spectral type , but the star appears too bright in @xmath66 .
it is likely that they are sirius - like systems in which the k star dominates the emission in longer wavelength and the white dwarf dominates the emission in shorter wavelengths .
the individual results of the stellar counterpart analysis are given in table [ addcount ] and a summary by category is provided in table [ countsum ] .
agns and galaxies can also be diagnosed from the seds by their distinct shape that rises toward long wavelengths in the infrared @xcite .
for example , sources # 11 , 24 , 36 , 38 , 45 , 63 , 64 , 98 , 117 , and 123 have well populated seds rising in the ir and are thus good candidates for agns .
indeed , seven of them ( # 11 , 38 , 45 , 64 , 98 , 117 , and 123 ) have been identified to be quasars by @xcite and one ( # 24 ) is a resolved galaxy with a prominent nucleus @xcite . since stellar emission is not expected to be flat and since many of these objects are detected only in the ir ( not in optical ) , we suggest that objects with flat seds may also be agns . in table
[ addcount ] , we use `` agn '' to denote confirmed objects and `` ( agn ) '' for candidates . finally , our observation did not reveal active hmxbs amongst known massive stars in n11 ( see sect . 4 ) and
the late b - stars newly identified with the photometry ( see previous paragraphs ) display no definitive sign of being in hmxbs .
we have further examined the data to see whether other bright x - ray sources would have ( unknown ) ob counterparts , following the method of @xcite and using the information available on the nature of the sources , when existing . as in 30
dor @xcite and some other lmc clusters @xcite , no hmxb was detected in n11 , although its cluster lh9 is old enough ( @xmath67myr , @xcite ) for its initially most massive stars to have undergone supernovae . adopting the classical salpeter initial mass function ( imf ) for the observed stellar content and with upper cut - off at @xmath68 , up to 10 stars initially more massive than @xmath69
might have been present in the cluster .
however , the production efficiency of hmxbs is low , which could explain the absence of detections in n11 ( e.g. , * ? ? ?
* ; * ? ? ?
* though not necessarily in agreement with galactic census of hmxbs from @xcite ) .
of the four massive star clusters in n11 , two ( lh9 and 10 ) are covered by our _ chandra _ observation .
their stellar content is quite well known . the first complete study ( * ? ? ?
* hereafter pgmw ) provided spectroscopic classification for about 75 stars , among which 43 are o - type objects . using _ hst _ data to disentangle the two compact ob groups hd32228 in lh9 and pgmw3204/9 in lh10 , @xcite gave spectral classification for 20 additional hot stars , while the vlt flames survey led to the discovery of 25 additional o - type stars , including a potential runaway star of type o2.5iii ( @xcite , hereafter elst ) .
in total , 1 wolf - rayet , 81 o - type stars and 80 b - type stars are now known in the fov .
the spectral monitoring of elst further indicated a binary fraction of 36% amongst massive stars of lh9 - 10 , which is quite low for such stars @xcite .
finally , @xcite performed atmosphere modelling on 22 of the elst targets , leading to the first accurate determination of their physical properties .
these studies showed that lh9 and lh10 appear very different from each other . indeed , lh9 is the most extended and the richest cluster in n11 .
it is dominated by a compact group of stars collectively referred to as hd32228 , which contains one wolf - rayet star of the carbon sequence ( brey 9 ) and many late o - type stars @xcite .
furthermore , stars in lh9 have blown a large superbubble @xcite .
in contrast , the lh10 cluster , still partly embedded in its natal cloud , appears rather young , with wind - blown bubbles of limited size @xcite and earliest o - type stars still present @xcite .
the imf in these clusters was constrained to @xmath70 and @xmath71 ( pgmw ) , confirming that there are many more high - mass stars in lh10 than in lh9 .
these facts , together with the higher reddening of lh10 , indicate the relative youth of lh10 compared to lh9 ( about 2myr difference , see pgmw and @xcite ) . because of the age difference and of the position of lh10 at the periphery of the lh9 superbubble
, it was suggested that the pair constitutes an example of sequential star formation @xcite .
the poorer psfs of the previous x - ray observations of n11 ( _ rosat _ - @xcite ; _ xmm - newton _
- @xcite ) did not lead to unambiguous detections of x - ray emission associated with the ob stars of lh9 or lh10 . at first sight ( see fig . [
fig : x14_sou ] ) , the _ chandra _ x - ray point sources do not appear obviously clustered towards lh9 , lh10 or their periphery , due to the large number of background sources ( sect .
however , correlating the list of x - ray sources ( see table[sourcelist ] ) with the list of ob stars with known spectroscopic classification led to 13 positive matches . as a last check , we further inspected the x - ray image by eye and one additional x - ray source associated with an o - type star was clearly spotted ( see e.g. the many countours showing a source associated with pgmw3100 in fig .
[ lh10hst ] ) .
an extraction run specific to ob stars ( see below ) yields a 2@xmath13 detection for this object and a false detection probability just below that of the detected objects ( i.e. much higher than for the truly undetected o - stars ) .
it lies near a rather bright point source and on the edge of diffuse emission , which probably explains its non - detection by the automated detection algorithms .
table[deto ] lists the properties of these 14 ob stars with detected x - ray emission .
the first five columns report the x - ray source number , the elst or pgmw identification , the spectral type and binary status ( from elst , @xcite , or pgmw ) .
in addition , we derived for each object the color excesses @xmath72 from the @xmath73 photometry , using the intrinsic colors from @xcite .
galactic reddening towards the lmc was estimated to be 00.15 mag by @xcite ; as the reddening of some of our stars is @xmath74 , we will consider the galactic contribution negligible and we therefore calculate the absorbing columns @xmath75 using a gas - to - dust ratio of @xmath76h - atom@xmath7mag@xmath1 , typical of the lmc @xcite .
the bolometric luminosities , when not estimated by @xcite , were also derived from the @xmath73 photometry with intrinsic colors and bolometric corrections from @xcite ; they were derived using the properties of the earliest component if the object is a binary with types of the two components known ( ` sb2 ' in table[deto ] ) .
these derived absorbing columns and bolometric luminosities are presented in the sixth and seventh columns of table[deto ] .
the last two columns of that table give the unabsorbed flux ( in the 0.510kev band ) and ratio ( see also fig .
[ lxlbol ] ) .
these x - ray luminosities were derived by different methods depending on the source brightness .
five of the sources have enough counts ( @xmath7780 cts ) for a rough spectral analysis .
they were modeled within xspec using an absorbed thermal model ( xspec models mekal and vphabs , both with metal abundances set to 0.3 times solar ) . results of these fits are shown in table[speco ] . as is usual for massive stars ( e.g. , * ? ? ?
* ) , two solutions of rather similar quality can be found , one with high t and low @xmath78 , the other with low t and high @xmath78 .
it is difficult to choose between the two solutions since the exact amount of additional absorption @xmath78 , due to the cool wind , beyond the interstellar absorption can not be a priori fixed : the formal best - fit one is listed in table[speco ] .
however , it must be stressed that this dual solution ambiguity affects only the derivation of intrinsic emission levels , but has no impact on the derivation of ratios , as the x - ray luminosities used in this context are corrected by the _
interstellar _ absorption only . for the fainter sources
, we derived fluxes by converting the count rates . using the on - axis response matrices and a thermal model with a temperature @xmath79=0.6kev absorbed by the @xmath75 determined above ( both with metal abundances set to 0.3 times solar ) , we derived conversion factors between count rates ( in cts ks@xmath1 ) and unabsorbed fluxes ( in erg@xmath7s@xmath1 ) of 0.52@xmath80 .
using a temperature of 0.3kev would increase the fluxes by less than a factor of 2 .
only one of the detected sources is a b - type star , the detection fraction for these objects is thus low , @xmath51.3% ( 1 out of 80 ) .
this is not surprising since b - type stars are generally not bright x - ray emitters and the x - ray emission apparently associated with such objects has generally been attributed to a flaring pms companion ( see e.g. , * ? ? ?
* ; * ? ? ?
* ) or to an accreting compact companion . generally , the b stars with detected x - ray emission have early types ( b0-b0.5 ) and their emission is due to embedded wind - shocks as in o - type stars .
this could apply to src # 110 ( elst33 ) , an early - type b star whose count rate is correspondingly low with a signal - to - noise ratio of @xmath81 and whose is similar to that of o - stars in our sample .
most of the detected sources correspond to hot , o - type objects .
two of them correspond to the compact ob groups hd32228 and pgmw3204/9 , while another one is associated with the compact region n11a and its small cluster ( see individual discussion for these below ) . excluding the 19 stars in these compact groups , the fraction of detected o - type stars is 16% ( 10 out of 62 ) .
when ranked by magnitude or spectral types , it appears that the hottest and/or earliest objects are preferentially detected , as could be expected ( the detection fraction is 57% for o25 stars , 25% for o5.5o7 stars and
3% for o7.59.7 stars ) .
however , when one looks into details , things are not so simple . for example , of the four o23 stars present in the field , only two are detected : the one belonging to the pgmw3204/9 compound ( which thus contains additional massive stars emitting x - rays ) and elst31 .
the latter object is neither a known binary nor the brightest earliest - type star ( elst26 being slightly brighter than elst31 ) .
to gain further insight into the global properties of the ob star population , we estimated the total - band count rates of all ob sources in the field using their catalogued positions and corresponding 90% eer ( i.e. not 70% eer as before ) to minimize the potential effect of the astrometry errors in the x - ray data and the objects positions .
results for the 13 detected o - stars are of course consistent with the count rates reported in table [ sourcelist ] . on the other hand , for the 52 undetected o - stars , these measurements allowed us to derive upper limits , which we adopted to be the 3@xmath13 errors on the count rates .
these upper limits on the count rates were transformed into upper limits on the unabsorbed flux as was done above for the fainter detected sources .
table[allo ] summarizes the properties of these 52 undetected o - type stars present in the fov .
columns one and two give the star numbers in elst and pgmw , respectively ; the third and fourth columns indicate the spectral type and binary status ; the fifth column provides the bolometric luminosity as given by @xcite when available or as derived from the @xmath73 photometry otherwise ; the sixth column lists the absorbing columns , derived from the photometric color excess as explained earlier ; the last columns yield the upper limits on the x - ray luminosity and on the ratio ( see also fig.[lxlbol ] ) .
in addition to these limits , we also derived the combined properties of these 52 undetected o - stars .
summing their count rates and combining the associated errors yields a stacked value of @xmath82ctsks@xmath1 , corresponding to an x - ray luminosity of @xmath83ergs@xmath1 and a of @xmath84 .
this value is expectedly lower than all individual lower limits on .
it also implies that several undetected massive stars ( since the stacked signal appears dominated by @xmath8510 objects ) emit with levels not much below our sensitivity limit , but only more sensitive observations with a high spatial resolution will be able to clearly detect them above the bright diffuse background .
having all that information at hand , we may now compare n11 with the galaxy . in our galaxy , massive o - stars display soft spectra ( @xmath79 of 0.20.6kev ) and follow @xmath86 @xcite .
this relation reflects an intimate link of x - rays and the stellar winds , as x - rays are produced behind shocks in these unstable outflows , though a full understanding of the origin of the relation is still lacking ( steps in this direction have been made by * ? ? ?
* ; * ? ? ?
different values have been reported in the literature for galactic objects , but these differences are probably explained by choices made in the analysis ( e.g. , method for deriving @xmath87 ) and the data quality @xcite . since the x - ray emission depends on the winds , it may be expected that massive stars with different wind properties , such as low - metallicity , will show a different level of x - ray emission .
in addition , deviations from that ` canonical ' @xmath88 relation are also found in exceptional cases .
for example , while most massive o+ob binaries are not much harder or brighter ( small overluminosities of 0.2dex at most , @xcite ) than single stars , a few systems appear hard and overluminous because an x - ray - bright wind - wind collision is present ( e.g. , hd93403 , @xcite ; cygob2#9 , @xcite ) .
magnetically confined winds may also lead to overluminosities in strongly magnetic stars ( e.g. , @xmath89oric , @xcite ) . in n11
, results from spectral fits suggest rather large ratios and high plasma temperatures .
indeed , except for the hd32228 compound , all detected sources clearly lie above the galactic @xmath90 relation ( fig.[lxlbol ] ) and temperatures of 0.71.3kev are recorded here , when temperatures of 0.2 or 0.6kev ( depending of the trade - off between temperature and absorption mentioned above ) are usually observed for galactic o - stars . while it is possible that the stellar census is limited by confusion ( with several neighbouring stars mistaken as one , leading to errors in the bolometric luminosity estimates ) , it is also very probable that we detect only the x - ray - bright tip of the massive star population .
the detected objects would then be wind - wind interacting systems or strongly magnetic objects ; they would therefore not be fully representative of the properties of the o - type population in n11 . in this context
, it may be worth noting that , excluding the stars in compact groups , only 20% ( 2 out of 10 ) of the x - ray sources are known binaries , i.e. , a smaller fraction than in the n11 population ( 36% , elst ) - though the limited monitoring of elst may have missed the multiplicity of some objects .
further study should be undertaken to clarify the status of the detected objects ( bona - fide single stars , colliding - wind binaries , magnetic objects ) . to get a more representative idea of the actual x - ray emission level from massive stars in the lmc
, we may turn to the undetected o - stars : their stacked emission suggests a lower value of @xmath91 . this is a value comparable to galactic values ( e.g. , carina nebula , @xcite ) , which contradicts a priori intuitions ( lower metallicity@xmath92weaker winds@xmath92fainter x - ray emission ) .
future observations are however needed to confirm the overall representativity of this value for massive stars , and to enlarge the study of the ratio in the lmc .
the main component of lh9 , the compact group also known as hd32228 , is clearly detected in our observations ( fig .
[ lh9 ha ] ) .
its x - ray emission follows the visible one , i.e. , it is not a simple point source and looks rather extended .
the morphology is actually reminiscent of a point - like source superimposed on a small region of diffuse emission , itself immersed in the fainter superbubble emission from the whole cluster .
we have extracted spectra of each of these 3 regions ( see tables [ speco ] and [ specdif ] ) .
the emission peaks at the position of the stellar group encompassing the wolf - rayet brey 9 or bat99 - 10 .
since this evolved object belongs to the carbon sequence , it is not expected to emit significant amounts of x - rays if single @xcite . the total luminosity from all o - type stars in the hd32228
compound amounts to @xmath93ergs@xmath1 , implying a ratio of @xmath94 , fully compatible with the galactic value ( table [ deto ] ) .
the surrounding area ( ` near hd32228 ' , see table[specdif ] ) , once cleaned of the superbubble ` background ' contribution , presents a harder spectrum than the emission associated with the superbubble itself .
its high temperature and high absorption suggests the emission to be dominated by the overall x - rays from unresolved stellar objects , most probably o and pms stars .
in contrast , another region in lh9 with higher surface brightness ( ` inside lh9 ' , see table[specdif ] ) displays a spectrum very similar to the superbubble itself - only a few deviant high - energy bins , maybe spurious , are detected as a very high temperature plasma with low emission measure . to the limits of our data , the x - ray emission associated with the superbubble around lh9 thus appears rather uniform if one excludes hd32228 and its surroundings .
[ lh10hst ] shows a close - up on the lh10 cluster . in this region ,
six point - like sources are detected .
they correspond to five o - type stars ( elst 31 , 38 , and 50 , pgmw 3070 and 3120 ) and one compact group ( pgmw3204/9 ) .
the brightest source , pgmw3070 , appears actually as a tight cluster in the _ hst _ images .
the point - like source associated with the two subclusters display a slightly elongated shape , probably indicating that they are not truly point - like sources .
their flux is also too high for a single star , with of @xmath95 for the sole pgmw3070 .
the fluxes of the other sources in lh10 are also higher than observed in galactic single o - type stars .
this is especially the case of the bright binary elst50 , which presents the highest ratio : even taking into account the presence of two stars , the x - ray source is still about 5 times brighter than expected .
the ratio appears closer to the galactic value only for pgmw3204/9 , considering the sum of the individual luminosities of this stellar group s components detailed stellar content is well known in this area , but maybe not elsewhere in lh10 .
it is also interesting to note that in our data , pgmw3204/9 appears as bright as pgmw3070 , whereas it was half as bright during the _ xmm _ observations ( compare our fig.[lh10hst ] with fig .
11 of @xcite ) : this may indicate some variability in its high - energy emission , which is not expected for single , ` normal ' massive stars and is therefore generally associated with an orbital or rotational modulation .
further monitoring of these objects is needed to ascertain their nature .
some diffuse emission is also present throughout the field , especially to the southwest of the cluster , near pgmw 3070 and 3120 . in this region ,
a wind - blown bubble was detected by @xcite .
however , the same authors found another expanding bubble east of pgmw 3204/9 and elst 50 , with similar expansion velocity but it does not appear to be associated with diffuse emission as bright as the former .
higher velocity structures were also detected by @xcite , but appear to the south and west of the x - ray bright region , not coincident with it . it might be noted that the diffuse emission , which appears centrally peaked rather than bright - rimmed , corresponds to the region of highest stellar density - the eastern part of lh10 being much less crowded .
this could suggest that at least part of the x - ray emission actually comes from unresolved stars , as in the case of ngc602 in the smc @xcite .
this hypothesis seems confirmed when the spectrum of that diffuse emission is analyzed ( table[specdif ] ) : two thermal components are present , one with a low temperature typical of soft diffuse emission and one with higher temperature , most probably stellar in origin ( both components have similar luminosities in the 0.32.0kev band ) .
the compact hii region n11a , lying to the east of lh10 , probably harbors the youngest optically - visible stars in the field @xcite .
it is associated with the ` star ' pgmw3264 or elst28 , which is actually the earliest component of a compact stellar group composed of 7 objects .
the associated x - ray emission is rather strong and a comparison of its shape with that of its neighbours suggests some extension . since the earliest object was proposed to be a highly obscured mid o - type star @xcite , so that strong x - ray emission from that star is not expected .
the x - ray emission may rather be associated with wind - wind interactions between the cluster members or a combination of several unresolved sources .
unfortunately , the spectrum of this source is difficult to extract since it lies on the edge of some datasets .
the crude x - ray spectrum only points to a high temperature , but the high noise prevents us to draw definitive conclusions .
these results should therefore be taken with caution ( especially since a low absorbing column is also favored by the noisy data , clearly at odds with the heavy extinction expected within a dense cloud ) .
n11a deserves further investigation especially to assess its stellar content in detail . as can be seen in the color image of the fov ( fig.[fig : rgb ] )
, soft diffuse emission pervades the n11 region . on the basis of the surface brightness
, we defined 11 regions to be further analyzed and three regions displaying lower emission were chosen as backgrounds ( see fig.[difpos ] and table[regiondif ] for their positions ) . for their study , we need to estimate the non - x - ray background contribution which is not vignetted by the telescope .
this estimate uses the acis stowed background database , which has been processed with charge transfer inefficiency ( cti ) and gain corrections . the background level in each chip of individual observations
was further normalized according to the ratio of its count rate to that of the stowed data in the 1012kev band , where events are almost completely due to the non - x - ray background .
spectra of the chosen regions were extracted , the non - x - ray background being subtracted from each one .
regions of twice the 70% eer around each point source were removed .
note that for visual presentation of smoothed diffuse x - ray intensity maps , we replace the source - removed region with values interpolated from data in surrounding bins ( see e.g. fig.[difpos ] ) .
the spectra have been analyzed in the 0.32kev range ; beyond which there is little signal . to take into account the variations of the sensitivity across the fov , the spectral
fitting was done in two steps .
first , each background ( or combination of background ) was individually fitted although the background spectral shape remains rather similar , with the most varying parameter being the overall intensity level .
second , the sources were fitted by models of the form @xmath96 , where the former component represents the true spectrum of the diffuse source ( metal abundances of the absorption and thermal emission models were accordingly fixed to 0.3 ) and the latter represents the best - fit background model determined before ( these were fixed to the best - fit background model with the normalisation factors scaled by the effective surface ratio - keyword bacscal - of the source and background regions ) .
the results of these fits are presented in table[specdif ] .
the spectral analysis of two additional regions of lh9 that display a higher surface brightness is also presented in this table .
the results from spectral fitting allow us to compare the selected regions .
first , we examine temperature variations .
the spectra of all regions display a thermal component at low temperature ( @xmath970.2kev ) . for lh9 and lh10 , an additional , hotter plasma component ( @xmath98kev )
was needed to obtain a good fit , suggesting it to be associated with unresolved stellar objects ( pms , o stars ) .
this contamination is not negligible since it represents about one - third of the total flux associated with the diffuse sources .
its stellar origin is supported by the analysis of the region ( called ` middle ' ) situated between lh9lh10 and the northern limit of the fov .
the spectrum of the latter region does not require the presence of a high temperature component : indeed , only a few stars are scattered across this region and the stellar contribution is thus expected to be negligible here .
note that a fit of the lh9 spectrum where the abundance of the low - temperature component is allowed to vary does not improve the @xmath34 . regarding absorptions ,
a value of @xmath99h - atom@xmath7 is generally found for the 1 t fits , and a lower value for the 2 t fits .
the larger absorption value might in fact be an artifact : if we use only one component to fit the spectra of lh9 or lh10 , the best - fit also yields a higher absorption for these objects - this may also be linked to the temperature - absorption trade - off mentioned before .
unfortunately , the spectra of the other diffuse sources are too noisy for making a meaningful 2@xmath100 fit . keeping this caveat in mind
, one can however see the remarkable homogeneity of the derived spectral parameters , with the exception of harder components being needed when stellar clusters are located ( lh9 , lh10 ) .
this paper reports the first results of a very deep observation of the giant hii region n11 in the lmc .
soft diffusion emission is seen throughout the field , but its spectra reveal some point source contamination .
thanks to the long exposure ( @xmath5300ks ) and the high spatial resolution , 165 x - ray point - sources were detected in the field , with three showing significant temporal variability .
our observation thus increases by more than a factor of five the number of point - sources known in n11 . keeping in mind that the sensitivity varies across the field
, it must be noted that the faintest detected sources have count rates of about 0.04 ctsks@xmath1 , which correspond to luminosities of about 10@xmath0ergs@xmath1 in the 0.58.0kev energy band .
diffuse emission is also detected throughout the field , but the harder x - ray emission from some regions indicates contamination from unresolved stars .
most of the x - ray sources are background objects seen through the lmc , but there are also 11 galactic stars .
however , 14 ob stars are clearly detected in x - rays , three of them corresponding to compact clusters ( hd32228 , pgmw3204/9 , and n11a ) .
the known binaries are not preferentially detected in n11 , though this conclusion might be biased by the incomplete knowledge of the stellar multiplicity .
indeed , these stars could correspond to interacting winds systems or magnetic objects , as may be suggested by their rather high luminosities and plasma temperatures . in this context , it should be noted that changes are detected in massive stars of lh10 , compared to older _ xmm _ data .
follow - up observations are therefore needed to ascertain the nature of these sources .
the stacked emission of the undetected o - stars yields a of @xmath84 .
this suggests that the intrinsic x - ray emission of massive stars could be similar in the galaxy and the low - metallicity environment of n11 .
this is unexpected , as x - ray emission from massive stars is known to arise in their line - driven stellar winds whose properties are known to vary with metallicity .
further observations are however needed to confirm this result .
yn acknowledges support from the fonds national de la recherche scientifique ( frs - fnrs , belgium ) , the university of lige ( incl .
arc ) , the university of illinois at urbana - champaign , and the prodex xmm and integral contracts .
yhc and rg acknowledge grant sao g07 - 8091a , qdw grant sao g07 - 8091b and lmo support from bmwi / dlr grant fkz 50 or 1302 . ; the subtracted background is estimated locally in csmooth .
_ right : _
h@xmath33 image for comparison .
the main regions are labelled . ,
title="fig:",height=288 ] ; the subtracted background is estimated locally in csmooth .
_ right : _
h@xmath33 image for comparison .
the main regions are labelled . , title="fig:",height=288 ]
uncertainties , are @xmath850.2 , see table[sourcelist ] ) .
sources are indicated by different symbols according to their nature ( see table [ addcount ] and top of figure ) - note that foreground stars may be active binaries , explaining their high hrs . for clarity ,
errors are not shown for the sources whose nature is uncertain . also included in the plot are hardness - ratio models derived from pimms : the solid thick curves are for the power - law model with photon index equal to 3 , 2 , and 1 , whereas the solid thin curves are for the thermal plasma with abundance of 0.4 times solar and temperatures equal to 0.3 , 1 , and 4 kev , from left to right . the equivalent hydrogen absorbing column densities @xmath75 are 1 , 10 , 50 , 100 , and @xmath101 ( dashed curves from bottom to top ) .
the lowest values correspond to absorptions of sources in the galaxy or lmc , while the largest values are representative of obscured sources behind lmc .
note that power laws models are just examples , but they may not be representative of the spectral properties of all displayed sources . , height=288 ]
sources are indicated by different symbols according to their nature ( see table [ addcount ] ) .
the main sequence magnitudes and colors are taken from @xcite for o - stars ( masses @xmath102m@xmath103 ) and @xcite for the other spectral types ( masses in the range 0.115m@xmath103 ) ; it was shifted by @xmath104 for the color - magnitude diagram .
the dotted line shows the intrinsic ( i.e. , dereddened ) colors of classical t tauri stars @xcite , the dashed lines correspond to increasing values of absorption ( using @xmath105 and @xcite extinction law ) for the blue and red limits of the main sequence for low - mass stars and of the tts sequence .
[ hr],title="fig:",height=288 ] photometry .
sources are indicated by different symbols according to their nature ( see table [ addcount ] ) .
the main sequence magnitudes and colors are taken from @xcite for o - stars ( masses @xmath102m@xmath103 ) and @xcite for the other spectral types ( masses in the range 0.115m@xmath103 ) ; it was shifted by @xmath104 for the color - magnitude diagram .
the dotted line shows the intrinsic ( i.e. , dereddened ) colors of classical t tauri stars @xcite , the dashed lines correspond to increasing values of absorption ( using @xmath105 and @xcite extinction law ) for the blue and red limits of the main sequence for low - mass stars and of the tts sequence .
[ hr],title="fig:",height=288 ] relation for o - type stars in the lmc as derived from the data of n11 .
the three asterisks and the ten filled circles correspond to the compact groups and other detected o stars , respectively ( see table[deto ] for details ) , whereas the arrows give the upper limits on the x - ray luminosity for the undetected objects ( see table[allo ] for details ) .
the solid line indicates @xmath106 , the dotted line is 0.5dex higher .
[ lxlbol ] ] image ( right , from * ? ? ?
* ) with x - ray contours superimposed ( at levels of 0.8,1.5,2,5 ctsks@xmath1arcmin@xmath107 ) .
the detected massive stars are labelled by their x - ray source number along with their star name between parentheses .
[ lh10hst ] ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) . [ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) . [ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] for fits ) and of the diffuse sources ( see table [ specdif ] for fits ) .
[ specima],title="fig : " ] lcccccc & r.a .
( j2000 ) & dec .
( j2000 ) & exp & roll ang & obs date + obsid & ( h m s ) & ( @xmath108)&(s ) & ( @xmath11 ) & ( yyyy - mm - dd ) + 8210 & 04 55 50.6 & -66 26 42 & 45486 & 187.7 & 2007 - 05 - 29 + 8469 & 04 55 50.2 & -66 26 34 & 48895 & 170.1 & 2007 - 06 - 17 + 8468 & 04 55 50.2 & -66 26 31 & 47359 & 165.1 & 2007 - 06 - 22 + 8467 & 04 55 50.2 & -66 26 29 & 48383 & 160.1 & 2007 - 06 - 27 + 8470 & 04 55 50.4 & -66 26 21 & 42492 & 143.8 & 2007 - 07 - 14 + 8211 & 04 55 50.9 & -66 26 14 & 47359 & 130.2 & 2007 - 07 - 31 + lrrrrrrrrrrrr 1 & j045427.48 - 662518.0 & 1.8 & @xmath109 & @xmath110 & @xmath111 & 8.4 & 183 & 42.6 & 191 & t + 2 & j045434.92 - 662602.1 & 1.5 & @xmath112 & @xmath113 & @xmath114 & 7.6 & 210 & 33.6 & 217 & t + 3 & j045438.74 - 662207.1 & 2.4 & @xmath115 & & & 8.5 & 77 & 44.8 & 223 & t + 4 & j045440.30 - 662408.4 & 1.5 & @xmath116 & @xmath117 & @xmath118 & 7.5 & 251 & 35.7 & 178 & t + 5 & j045442.40 - 662227.4 & 1.7 & @xmath119 & @xmath120 & @xmath121 & 8.0 & 193 & 40.3 & 218 & t + 6 & j045443.47 - 662518.9 & 1.3 & @xmath122 & @xmath123 & @xmath124 & 6.9 & 121 & 24.1 & 208 & t + 7 & j045444.36 - 662551.0 & 1.2 & @xmath125 & @xmath126 & @xmath127 & 6.7 & 213 & 27.7 & 144 & s + 8 & j045444.89 - 662957.1 & 1.6 & @xmath128 & @xmath129 & & 7.3 & 96 & 32.3 & 221 & t + 9 & j045446.22 - 663146.7 & 1.9 & @xmath130 & & @xmath131 & 8.2 & 95 & 31.5 & 117 & t + 10 & j045446.51 - 662543.0 & 1.2 & @xmath132 & @xmath133 & @xmath134 & 6.5 & 205 & 60.8 & 158 & s + 11 & j045446.60 - 662004.4 & 2.1 & @xmath135 & @xmath136 & @xmath137 & 9.2 & 633 & 37.5 & 133 & t + 12 & j045447.01 - 662600.9 & 1.2 & @xmath138 & & @xmath139 & 6.4 & 106 & 51.5 & 139 & s + 13 & j045447.23 - 662759.2 & 1.1 & @xmath140 & @xmath141 & @xmath142 & 6.5 & 256 & 23.5 & 205 & t + 14 & j045447.36 - 662605.7 & 1.2 & @xmath143 & & @xmath144 & 6.3 & 91 & 43.4 & 139 & t + 15 & j045448.38 - 662103.3 & 1.8 & @xmath145 & @xmath146 & @xmath147 & 8.4 & 199 & 38.1 & 193 & t + 16 & j045448.45 - 662535.9 & 1.1 & @xmath148 & & @xmath149 & 6.3 & 166 & 55.9 & 149 & s + 17 & j045449.00 - 663120.4 & 1.7 & @xmath150 & @xmath151 & & 7.7 & 96 & 34.8 & 165 & t + 18 & j045449.85 - 662504.9 & 1.3 & @xmath152 & & & 6.3 & 66 & 37.0 & 155 & t + 19 & j045451.45 - 662530.0 & 1.1 & @xmath153 & & @xmath154 & 6.0 & 92 & 31.8 & 131 & s + 20 & j045451.80 - 662600.1 & 1.1 & @xmath155 & & @xmath156 & 5.9 & 59 & 22.3 & 157 & s + 21 & j045452.49 - 662644.3 & 1.0 & @xmath157 & @xmath158 & @xmath159 & 5.8 & 455 & 14.8 & 186 & t + 22 & j045452.85 - 662557.8 & 1.1 & @xmath160 & & @xmath161 & 5.8 & 39 & 12.8 & 151 & s + 23 & j045454.26 - 663015.4 & 1.3 & @xmath162 & @xmath163 & @xmath164 & 6.6 & 119 & 22.8 & 216 & t + 24 & j045454.76 - 662704.0 & 1.0 & @xmath165 & @xmath166 & & 5.6 & 60 & 16.3 & 220 & t + 25 & j045458.44 - 662240.8 & 1.5 & @xmath167 & & & 6.6 & 64 & 28.5 & 217 & t + 26 & j045501.67 - 663348.1 & 2.3 & @xmath168 & & & 8.6 & 37 & 9.4 & 32 & t + 27 & j045501.86 - 663147.2 & 1.3 & @xmath169 & @xmath170 & @xmath171 & 7.0 & 253 & 30.4 & 211 & t + 28 & j045501.90 - 663122.8 & 1.2 & @xmath172 & @xmath173 & @xmath174 & 6.7 & 172 & 8.1 & 146 & s + 29 & j045502.08 - 662300.4 & 1.4 & @xmath175 & & & 6.1 & 32 & 11.8 & 251 & h + 30 & j045502.41 - 662804.8 & 1.0 & @xmath176 & & & 5.0 & 33 & 12.1 & 221 & t + 31 & j045508.06 - 662139.6 & 1.5 & @xmath177 & & & 6.6 & 55 & 25.8 & 211 & t + 32 & j045509.20 - 663018.5 & 0.9 & @xmath178 & @xmath179 & @xmath180 & 5.5 & 418 & 13.2 & 229 & t + 33 & j045509.97 - 663318.3 & 1.8 & @xmath181 & & & 7.7 & 37 & 6.9 & 119 & s + 34 & j045513.36 - 663226.1 & 1.3 & @xmath182 & @xmath183 & & 6.8 & 176 & 27.5 & 237 & t + 35 & j045515.02 - 663039.3 & 0.9 & @xmath184 & @xmath185 & @xmath186 & 5.3 & 176 & 14.4 & 210 & t + 36 & j045518.07 - 661735.4 & 2.3 & @xmath187 & @xmath188 & @xmath189 & 9.7 & 514 & 49.4 & 131 & t + 37 & j045518.65 - 662156.3 & 1.1 & @xmath190 & @xmath191 & & 5.7 & 68 & 14.8 & 241 & t + 38 & j045519.02 - 662701.2 & 0.6 & @xmath192 & & & 3.2 & 20 & 3.2 & 241 & t + 39 & j045519.11 - 662242.7 & 0.9 & @xmath193 & @xmath194 & & 5.1 & 63 & 10.6 & 245 & t + 40 & j045521.93 - 662956.2 & 1.1 & @xmath175 & & & 4.3 & 10 & 1.3 & 136 & s + 41 & j045523.45 - 663040.9 & 0.9 & @xmath195 & & & 4.8 & 44 & 8.4 & 223 & t + 42 & j045524.11 - 662154.7 & 1.5 & @xmath196 & & & 5.5 & 16 & 3.4 & 136 & s + 43 & j045524.12 - 662831.4 & 0.8 & @xmath197 & & & 3.2 & 17 & 3.5 & 240 & t + 44 & j045524.13 - 662535.5 & 0.4 & @xmath198 & @xmath199 & @xmath200 & 2.9 & 72 & 2.5 & 223 & t + 45 & j045524.46 - 662223.8 & 0.8 & @xmath201 & @xmath202 & @xmath203 & 5.0 & 285 & 9.5 & 224 & t + 46 & j045525.18 - 663006.6 & 0.7 & @xmath204 & & @xmath205 & 4.2 & 43 & 7.2 & 197 & t + 47 & j045526.41 - 662226.4 & 0.8 & @xmath206 & @xmath207 & @xmath208 & 4.9 & 136 & 9.7 & 217 & t + 48 & j045529.39 - 662311.8 & 0.8 & @xmath209 & & & 4.1 & 25 & 6.1 & 186 & t + 49 & j045530.93 - 662516.4 & 0.5 & @xmath192 & & & 2.4 & 14 & 1.5 & 184 & t + 50 & j045531.60 - 662639.7 & 0.3 & @xmath210 & @xmath211 & & 1.9 & 63 & 1.6 & 243 & t + 51 & j045532.90 - 662404.6 & 0.6 & @xmath212 & & & 3.2 & 26 & 3.6 & 231 & t + 52 & j045533.23 - 663245.8 & 1.3 & @xmath213 & & & 6.3 & 63 & 20.1 & 199 & t + 53 & j045533.29 - 662936.9 & 0.8 & @xmath214 & & & 3.4 & 16 & 3.8 & 237 & t + 54 & j045533.70 - 662933.3 & 0.6 & @xmath215 & & & 3.3 & 20 & 4.2 & 210 & t + 55 & j045533.84 - 663015.2 & 0.6 & @xmath216 & @xmath217 & @xmath218 & 3.9 & 91 & 4.6 & 177 & t + 56 & j045534.88 - 661526.0 & 3.6 & @xmath219 & & @xmath220 & 11.4 & 44 & 8.7 & 20 & s + 57 & j045535.48 - 662736.7 & 0.7 & @xmath221 & & & 1.8 & 11 & 1.5 & 214 & t + 58 & j045535.59 - 663104.3 & 1.1 & @xmath222 & & & 4.6 & 18 & 3.8 & 228 & h + 59 & j045536.75 - 662526.8 & 0.3 & @xmath223 & @xmath224 & @xmath225 & 1.9 & 99 & 1.2 & 174 & t + 60 & j045536.79 - 662653.3 & 0.5 & @xmath197 & & & 1.4 & 16 & 1.2 & 266 & t + 61 & j045537.76 - 662517.7 & 0.5 & @xmath197 & @xmath226 & & 1.9 & 14 & 1.1 & 215 & h + 62 & j045538.24 - 662537.3 & 0.7 & @xmath227 & & & 1.6 & 5 & 0.3 & 143 & s + 63 & j045538.99 - 662550.2 & 0.3 & @xmath228 & @xmath229 & @xmath230 & 1.4 & 247 & 1.3 & 226 & t + 64 & j045539.69 - 662959.5 & 0.5 & @xmath231 & @xmath232 & @xmath233 & 3.5 & 1000 & 4.4 & 189 & t + 65 & j045540.12 - 662013.1 & 1.4 & @xmath234 & @xmath235 & & 6.6 & 59 & 24.5 & 235 & t + 66 & j045540.68 - 662613.3 & 0.5 & @xmath236 & & & 1.1 & 22 & 1.3 & 241 & t + 67 & j045540.79 - 662820.1 & 0.6 & @xmath227 & & & 1.9 & 6 & 0.3 & 196 & s + 68 & j045542.24 - 662610.2 & 0.4 & @xmath237 & @xmath238 & & 1.0 & 31 & 1.1 & 235 & t + 69 & j045542.82 - 663114.5 & 0.7 & @xmath239 & @xmath240 & @xmath241 & 4.6 & 66 & 10.6 & 223 & t + 70 & j045543.08 - 663303.6 & 1.3 & @xmath242 & @xmath243 & & 6.4 & 73 & 23.6 & 243 & t + 71 & j045544.02 - 662453.1 & 2.6 & @xmath221 & & & 1.9 & 6 & 0.4 & 175 & s + 72 & j045549.27 - 662654.4 & 0.4 & @xmath244 & & & 0.2 & 34 & 0.9 & 185 & t + 73 & j045549.98 - 663349.1 & 1.4 & @xmath245 & @xmath246 & @xmath247 & 7.1 & 172 & 38.6 & 211 & t + 74 & j045551.14 - 662050.7 & 1.2 & @xmath248 & & & 5.9 & 53 & 17.2 & 250 & t + 75 & j045552.38 - 662850.9 & 0.5 & @xmath249 & & & 2.2 & 28 & 2.2 & 235 & t + 76 & j045553.76 - 662831.5 & 0.3 & @xmath250 & @xmath251 & @xmath252 & 1.8 & 126 & 1.7 & 256 & t + 77 & j045554.05 - 662459.2 & 1.0 & @xmath227 & & & 1.8 & 6 & 0.4 & 157 & s + 78 & j045554.19 - 662818.9 & 0.3 & @xmath253 & @xmath254 & @xmath255 & 1.6 & 113 & 1.9 & 232 & t + 79 & j045554.52 - 662954.6 & 0.6 & @xmath192 & & & 3.2 & 12 & 1.9 & 158 & s + 80 & j045554.56 - 661832.2 & 2.0 & @xmath256 & & & 8.2 & 89 & 44.5 & 213 & t + 81 & j045554.75 - 662714.7 & 0.2 & @xmath257 & @xmath258 & @xmath259 & 0.7 & 300 & 1.7 & 251 & t + 82 & j045554.87 - 661701.0 & 2.4 & @xmath260 & & @xmath261 & 9.7 & 87 & 11.6 & 66 & s + 83 & j045555.82 - 662709.8 & 2.3 & @xmath262 & & & 0.7 & 6 & 0.4 & 250 & s + 84 & j045556.65 - 662223.0 & 0.6 & @xmath263 & @xmath264 & @xmath265 & 4.4 & 173 & 7.4 & 193 & t + 85 & j045556.89 - 663238.7 & 1.1 & @xmath266 & @xmath267 & & 6.0 & 74 & 16.4 & 191 & t + 86 & j045557.19 - 663130.5 & 0.8 & @xmath268 & @xmath269 & @xmath270 & 4.8 & 93 & 8.6 & 210 & t + 87 & j045558.10 - 661856.1 & 1.8 & @xmath271 & & & 7.8 & 47 & 10.2 & 176 & s + 88 & j045601.44 - 662936.2 & 0.4 & @xmath272 & @xmath273 & @xmath274 & 3.1 & 137 & 3.7 & 230 & t + 89 & j045601.67 - 661901.8 & 1.9 & @xmath275 & & & 7.8 & 77 & 36.8 & 205 & t + 90 & j045602.02 - 662616.3 & 1.3 & @xmath276 & & & 1.2 & 6 & 0.3 & 126 & s + 91 & j045602.14 - 662643.8 & 0.7 & @xmath221 & & & 1.2 & 11 & 1.7 & 238 & t + 92 & j045602.31 - 662403.2 & 0.6 & @xmath227 & @xmath277 & & 2.9 & 15 & 2.5 & 290 & h + 93 & j045602.53 - 662738.7 & 1.6 & @xmath262 & & & 1.5 & 6 & 0.4 & 194 & s + 94 & j045604.12 - 662338.7 & 0.6 & @xmath215 & & & 3.3 & 22 & 4.9 & 228 & t + 95 & j045604.18 - 662126.4 & 1.2 & @xmath278 & & & 5.4 & 22 & 4.3 & 161 & s + 96 & j045604.40 - 662950.2 & 0.5 & @xmath249 & & @xmath279 & 3.4 & 32 & 4.9 & 238 & t + 97 & j045604.60 - 662358.0 & 0.7 & @xmath192 & & & 3.1 & 15 & 0.9 & 190 & s + 98 & j045604.89 - 662638.7 & 1.3 & @xmath221 & & & 1.4 & 7 & 0.4 & 208 & s + 99 & j045605.09 - 662807.5 & 0.4 & @xmath280 & & @xmath281 & 2.0 & 37 & 2.4 & 258 & t + 100 & j045605.21 - 662508.4 & 0.3 & @xmath282 & @xmath283 & @xmath284 & 2.1 & 128 & 2.1 & 253 & t + 101 & j045606.08 - 661537.6 & 4.6 & @xmath285 & & & 11.2 & 17 & 4.3 & 21 & s + 102 & j045606.31 - 662649.7 & 0.4 & @xmath286 & @xmath287 & @xmath288 & 1.6 & 68 & 1.3 & 231 & t + 103 & j045607.45 - 663331.4 & 1.6 & @xmath289 & & & 7.0 & 71 & 32.0 & 192 & t + 104 & j045608.01 - 662404.1 & 0.5 & @xmath290 & & & 3.2 & 36 & 3.3 & 243 & t + 105 & j045608.50 - 662608.6 & 0.4 & @xmath291 & @xmath292 & & 1.9 & 35 & 1.1 & 223 & h + 106 & j045609.29 - 663455.5 & 2.2 & @xmath293 & & & 8.4 & 34 & 12.0 & 144 & s + 107 & j045610.30 - 663046.6 & 0.9 & @xmath278 & & & 4.5 & 23 & 2.9 & 160 & s + 108 & j045610.72 - 661938.7 & 2.0 & @xmath294 & & & 7.3 & 30 & 10.6 & 211 & s + 109 & j045610.96 - 662359.5 & 0.7 & @xmath236 & & & 3.4 & 21 & 4.4 & 201 & t + 110 & j045611.09 - 662824.6 & 0.6 & @xmath262 & & & 2.7 & 9 & 1.2 & 213 & s + 111 & j045613.10 - 662123.9 & 1.2 & @xmath234 & & & 5.8 & 52 & 20.4 & 216 & t + 112 & j045615.13 - 662410.9 & 0.5 & @xmath295 & @xmath240 & @xmath296 & 3.5 & 55 & 4.1 & 226 & t + 113 & j045617.78 - 662604.0 & 0.4 & @xmath297 & @xmath298 & @xmath299 & 2.8 & 76 & 3.4 & 229 & t + 114 & j045618.12 - 662721.0 & 0.5 & @xmath300 & & & 2.8 & 38 & 3.1 & 226 & t + 115 & j045619.28 - 662702.3 & 0.7 & @xmath301 & & & 2.9 & 10 & 1.1 & 158 & s + 116 & j045619.49 - 662631.0 & 0.4 & @xmath302 & @xmath303 & @xmath304 & 2.9 & 219 & 2.7 & 198 & t + 117 & j045620.14 - 662718.3 & 0.4 & @xmath305 & @xmath306 & @xmath307 & 3.0 & 155 & 3.8 & 258 & t + 118 & j045620.37 - 663010.1 & 0.9 & @xmath308 & & & 4.6 & 34 & 9.7 & 239 & t + 119 & j045621.49 - 662446.1 & 0.9 & @xmath309 & & & 3.6 & 9 & 1.0 & 188 & s + 120 & j045622.48 - 662837.5 & 0.8 & @xmath309 & & & 3.7 & 20 & 5.2 & 251 & t + 121 & j045622.87 - 663106.8 & 1.2 & @xmath310 & & & 5.5 & 22 & 4.7 & 177 & s + 122 & j045623.18 - 662702.8 & 0.5 & @xmath311 & @xmath312 & & 3.3 & 37 & 5.0 & 253 & t + 123 & j045624.56 - 662206.7 & 0.9 & @xmath313 & @xmath314 & @xmath315 & 5.7 & 556 & 17.8 & 203 & t + 124 & j045625.33 - 663029.8 & 0.9 & @xmath165 & & @xmath316 & 5.1 & 35 & 4.6 & 182 & s + 125 & j045626.22 - 662922.8 & 0.7 & @xmath317 & @xmath318 & @xmath319 & 4.5 & 40 & 2.3 & 187 & s + 126 & j045629.66 - 662138.5 & 1.5 & @xmath278 & & & 6.4 & 23 & 6.8 & 164 & s + 127 & j045629.74 - 662702.7 & 0.6 & @xmath320 & @xmath321 & @xmath322 & 3.9 & 212 & 6.0 & 208 & t + 128 & j045629.82 - 662224.0 & 1.0 & @xmath323 & & @xmath324 & 5.8 & 82 & 20.3 & 213 & t + 129 & j045632.32 - 661905.4 & 2.1 & @xmath325 & @xmath326 & & 8.7 & 146 & 64.1 & 220 & t + 130 & j045632.41 - 663012.9 & 1.0 & @xmath327 & & @xmath328 & 5.5 & 53 & 10.8 & 175 & t + 131 & j045632.61 - 663506.2 & 2.8 & @xmath329 & & & 9.4 & 31 & 10.9 & 82 & s + 132 & j045632.69 - 663100.5 & 1.2 & @xmath204 & & & 6.0 & 55 & 17.7 & 208 & t + 133 & j045634.30 - 663035.3 & 1.1 & @xmath330 & @xmath331 & @xmath332 & 5.8 & 85 & 14.6 & 236 & t + 134 & j045634.66 - 662827.2 & 0.8 & @xmath333 & & @xmath334 & 4.7 & 44 & 9.9 & 188 & s + 135 & j045634.87 - 662623.1 & 0.7 & @xmath335 & @xmath336 & @xmath337 & 4.4 & 94 & 8.6 & 227 & t + 136 & j045636.28 - 663123.4 & 1.2 & @xmath338 & @xmath339 & @xmath340 & 6.5 & 119 & 22.0 & 203 & t + 137 & j045637.44 - 661829.9 & 2.3 & @xmath341 & @xmath342 & @xmath343 & 9.5 & 184 & 52.6 & 148 & t + 138 & j045639.99 - 663402.3 & 2.4 & @xmath344 & & & 8.8 & 81 & 34.2 & 119 & t + 139 & j045640.74 - 662232.1 & 1.6 & @xmath308 & @xmath345 & & 6.5 & 48 & 24.4 & 239 & t + 140 & j045640.74 - 662655.7 & 0.8 & @xmath346 & @xmath347 & @xmath348 & 5.0 & 119 & 10.2 & 212 & t + 141 & j045642.55 - 662518.1 & 1.1 & @xmath177 & & @xmath349 & 5.4 & 28 & 4.6 & 158 & s + 142 & j045643.20 - 662502.7 & 0.9 & @xmath350 & @xmath351 & @xmath352 & 5.5 & 240 & 18.9 & 176 & t + 143 & j045643.77 - 662055.4 & 1.7 & @xmath353 & @xmath354 & @xmath355 & 7.9 & 136 & 50.0 & 209 & t + 144 & j045644.58 - 662921.1 & 1.4 & @xmath294 & & & 6.0 & 40 & 18.5 & 196 & t + 145 & j045645.00 - 662858.9 & 1.4 & @xmath356 & & & 5.9 & 44 & 21.5 & 216 & t + 146 & j045646.70 - 662446.6 & 1.1 & @xmath357 & & @xmath358 & 5.9 & 78 & 27.0 & 197 & t + 147 & j045647.73 - 662752.5 & 1.1 & @xmath359 & @xmath360 & @xmath361 & 5.8 & 82 & 19.3 & 203 & t + 148 & j045648.22 - 662401.0 & 1.1 & @xmath362 & @xmath363 & @xmath364 & 6.4 & 188 & 23.2 & 234 & t + 149 & j045649.22 - 663042.3 & 1.6 & @xmath167 & & & 7.1 & 63 & 27.9 & 222 & t + 150 & j045651.35 - 661905.4 & 3.2 & @xmath365 & & & 9.7 & 34 & 12.3 & 123 & s + 151 & j045655.52 - 663231.2 & 2.7 & @xmath366 & & & 8.7 & 66 & 36.4 & 145 & t + 152 & j045658.75 - 662438.9 & 1.4 & @xmath367 & @xmath368 & @xmath369 & 7.1 & 136 & 33.0 & 175 & t + 153 & j045659.64 - 662415.3 & 1.8 & @xmath177 & & & 7.3 & 69 & 35.6 & 238 & t + 154 & j045700.03 - 662535.0 & 1.5 & @xmath167 & & & 7.0 & 66 & 26.6 & 249 & t + 155 & j045700.05 - 662113.5 & 2.3 & @xmath370 & & & 8.9 & 103 & 57.1 & 222 & t + 156 & j045700.53 - 662508.7 & 1.6 & @xmath371 & & & 7.2 & 73 & 28.7 & 183 & t + 157 & j045700.94 - 662357.7 & 1.9 & @xmath160 & & & 7.6 & 42 & 13.1 & 165 & s + 158 & j045702.07 - 662257.1 & 1.9 & @xmath372 & @xmath373 & & 8.1 & 107 & 48.1 & 254 & t + 159 & j045702.86 - 663042.3 & 1.9 & @xmath374 & & & 8.2 & 89 & 36.3 & 166 & t + 160 & j045704.01 - 662131.9 & 2.1 & @xmath375 & @xmath376 & & 9.0 & 208 & 58.2 & 225 & t + 161 & j045705.85 - 661928.4 & 3.3 & @xmath377 & & & 10.4 & 73 & 34.4 & 94 & t + 162 & j045706.90 - 662705.1 & 1.5 & @xmath375 & @xmath378 & @xmath379 & 7.6 & 158 & 30.2 & 191 & t + 163 & j045714.71 - 662226.7 & 2.3 & @xmath380 & @xmath381 & @xmath382 & 9.4 & 183 & 52.0 & 174 & t + 164 & j045716.45 - 662322.6 & 2.5 & @xmath383 & & & 9.2 & 44 & 14.6 & 163 & s + 165 & j045728.37 - 662232.0 & 3.9 & @xmath384 & & & 10.6 & 23 & 6.7 & 57 & s + lccccccccccccc 4 & & & & & & & & & 0.67 & 04544030 - 6624078 & & & + 5 & & & & & 0.79 & s1iw031883 & & & 0.10 & 04544240 - 6622276 & 0.25 & 0.78 & 0236 - 0039098 + 11 & & & & & 0.89 & s1iw033748 & & & 0.59 & 04544661 - 6620039 & 0.51 & 0.14 & 0236 - 0039136 + 13 & & & & & & & & & 0.48 & 04544726 - 6627588 & & & + 17 & & & & & & & & & 0.57 & 04544895 - 6631200 & & & + 21 & & & & & & & & & 0.76 & 04545253 - 6626436 & & & + 22 & & & & & & & & & & & & 0.86 & 0235 - 0043997 + 23 & & & & & & & & & & & 0.85 & & + 24 & 0.48 & 04545476 - 6627035 & 0.44 & j045454.69 - 662704.1 & 0.44 & s1iw027637 & & & 0.23 & 04545478 - 6627038 & 0.69 & 0.21 & 0235 - 0044040 + 25 & & & & & & & & & 0.47 & 04545852 - 6622407 & 0.58 & & + 28 & 0.72 & 04550187 - 6631221 & 0.62 & j045501.82 - 663122.5 & 0.73 & s1iw084526 & & & 0.69 & 04550188 - 6631222 & 0.62 & 0.39 & 0234 - 0051103 + 33 & & & & & & & & & 0.53 & 04551006 - 6633181 & 0.54 & & + 34 & & & & & & & & & 0.81 & 04551324 - 6632258 & & & + 36 & & & & & & & & & 0.40 & 04551802 - 6617353 & & & + 38 & & & & & & & & & 0.17 & 04551900 - 6627012 & & & + 40 & & & & & & & & & & & & 0.58 & 0235 - 0044409 + 42 & & & 0.79 & j045524.04 - 662154.1 & 0.99 & s1iw031973 & & & & & & 0.59 & 0236 - 0039555 + 43 & & & & & 0.22 & s1iw025569 & & & & & & & + 44 & & & & & & & & & 0.70 & 04552424 - 6625353 & 0.74 & & + 45 & & & & & & & & & 0.42 & 04552445 - 6622235 & & & + 46 & & & & & & & & & & & 0.93 & & + 48 * & 0.14 & 04552941 - 6623118 & 0.64 & j045529.41 - 662312.5 & 0.43 & s1iw000197 & & & 0.22 & 04552940 - 6623121 & 0.39 & 0.43 & 0236 - 0039600 + 50 & & & & & & & & & 0.62 & 04553154 - 6626393 & & & + 54 & & & & & 0.42 & s1iw146637 & & & 0.49 & 04553366 - 6629329 & 0.20 & & + 55 & & & & & & & & & 0.16 & 04553386 - 6630151 & 0.16 & & + 57 & & & & & & & & & 0.33 & 04553554 - 6627367 & & & + 61 & & & & & & & & & 0.45 & 04553778 - 6625173 & 0.82 & & + 63 & & & & & & & 0.37 & 15 ( haebe ) & 0.36 & 04553899 - 6625499 & & & + 64 & & & & & & & & & 0.35 & 04553970 - 6629592 & 0.32 & 0.99 & 0235 - 0044624 + 65 & & & & & & & & & 0.81 & 04554018 - 6620124 & & & + 66 & & & & & & & & & & & 0.92 & & + 70 & & & & & & & & & & & 0.79 & & + 71 & & & & & & & & & 0.82 & 04554390 - 6624535 & 0.79 & & + 77 * & 0.57 & 04555396 - 6624593 & 0.67 & j045553.97 - 662459.7 & 0.52 & s1iw114172 & 0.69 & 83 ( ob ) & 0.70 & 04555394 - 6624592 & 0.79 & 0.74 & 0235 - 0044790 + 78 & & & & & & & & & 0.66 & 04555412 - 6628184 & 0.58 & & + 85 & & & & & & & & & 0.94 & 04555683 - 6632379 & & & + 86 & & & & & & & & & 0.75 & 04555725 - 6631299 & 0.79 & & + 87 & 0.09 & 04555811 - 6618561 & & & & & & & 0.17 & 04555812 - 6618560 & 0.32 & 0.89 & 0236 - 0039834 + 93 & & & & & & & & & & & 0.77 & & + 95 & 0.53 & 04560410 - 6621266 & & & 0.87 & s1iw147534 & & & 0.28 & 04560415 - 6621266 & 0.24 & & + 96 & & & & & & & & & & & & 0.94 & 0235 - 0044909 + 97 * & 0.18 & 04560457 - 6623580 & & & 0.21 & s1iw114345 & 0.12 & 113 ( ob ) & 0.11 & 04560460 - 6623579 & 0.20 & 0.71 & 0236 - 0039905 + 98 & & & & & & & 0.33 & 31 ( haebe ) & 0.33 & 04560492 - 6626385 & & & + 104 & & & & & & & & & & & 0.87 & & + 106 & & & & & & & & & 0.12 & 04560930 - 6634554 & 0.04 & & + 107 & & & & & & & & & 0.77 & 04561034 - 6630459 & & & + 108 & & & & & & & & & 0.78 & 04561066 - 6619381 & 0.81 & 0.66 & 0236 - 0039961 + 109 & & & & & & & & & 0.38 & 04561099 - 6623592 & 0.54 & & + 110 * & 0.54 & 04561102 - 6628242 & 0.68 & j045610.98 - 662824.5 & 0.60 & s1iw113493 & 0.52 & 140 ( ob ) & 0.51 & 04561102 - 6628243 & 0.50 & 0.97 & 0235 - 0044979 + 112 & & & & & & & & & 0.29 & 04561509 - 6624111 & & & + 115 * & 0.55 & 04561931 - 6627018 & 0.29 & j045619.27 - 662702.1 & 0.61 & s1iw113790 & 0.56 & 170 ( ob ) & 0.52 & 04561932 - 6627019 & 0.77 & 0.55 & 0235 - 0045088 + 117 & & & & & & & & & 0.28 & 04562013 - 6627181 & & & + 118 & & & & & & & & & 1.00 & 04562052 - 6630097 & & & + 122 & & & & & & & & & & & 0.79 & & + 123 & & & & & & & & & 0.24 & 04562456 - 6622065 & & & + 126 * & 0.37 & 04562970 - 6621388 & 0.57 & j045629.58 - 662138.8 & 0.09 & s1iw114755 & 0.33 & 209 ( ob ) & 0.40 & 04562969 - 6621389 & 0.20 & 0.72 & 0236 - 0040096 + 130 & 0.42 & 04563242 - 6630124 & 0.27 & j045632.38 - 663012.7 & 0.38 & s1iw113097 & 0.41 & 217 ( ob ) & 0.52 & 04563242 - 6630124 & 0.52 & & + 134 * & 0.87 & 04563466 - 6628263 & & & & & & & 0.20 & 04563469 - 6628271 & & & + 136 & 0.59 & 04563632 - 6631228 & 0.34 & j045636.28 - 663123.1 & & & 0.58 & 247 ( ob ) & 0.57 & 04563632 - 6631229 & 0.69 & & + 137 & & & & & & & & & & & & 0.30 & 0236 - 0040143 + 139 & & & & & & & & & 0.79 & 04564085 - 6622317 & 0.71 & & + 141 * & 0.57 & 04564246 - 6625180 & 0.44 & j045642.50 - 662518.4 & & & 0.28 & 278 ( ob ) & 0.28 & 04564251 - 6625181 & 0.42 & & + 142 * & 0.16 & 04564320 - 6625025 & 0.12 & j045643.18 - 662502.7 & 0.68 & s1iw000213 & & & 0.48 & 04564327 - 6625025 & 0.46 & 0.67 & 0235 - 0045338 + 145 & & & & & 0.89 & s1iw146732 & & & 0.49 & 04564492 - 6628590 & 0.52 & & + 146 * & 0.62 & 04564681 - 6624467 & 0.40 & j045646.77 - 662446.8 & & & & & 0.60 & 04564681 - 6624467 & 0.57 & & + 152 * & & & & & & & & & 0.93 & 04565882 - 6624381 & & & + 156 & & & & & & & & & 0.51 & 04570059 - 6625092 & & & + 157 * & 0.70 & 04570087 - 6623572 & 0.63 & j045700.86 - 662357.4 & & & & & 0.69 & 04570088 - 6623572 & 0.77 & & + 159 & & & & & & & & & 0.50 & 04570294 - 6630421 & 0.45 & & + 160 & & & & & & & & & 0.58 & 04570408 - 6621324 & 0.57 & & + 164 * & & & & & & & & & 0.38 & 04571651 - 6623225 & & & + lcccccll 1 & y & & & & & point source + diffuse x - ray & galaxy at z=0.024 + 4 & y & y & y & 0.56 & 045440.29 - 662407.9&sed rises in ir & ( agn ) + 5 & y & y & & 0.24 & 045442.38 - 662227.6 & & galactic k0v star at 4.5 kpc + 7 & & & & & & & n11l snr + 10 & & & & & & & n11l snr + 11 & y & y & y & 0.39 & 045446.61 - 662004.1 & & agn ( see sect .
+ 12 & & & & & & & n11l snr + 13 & & y & y & 0.27 & 045447.28 - 662759.2&sed peaking at 3@xmath48 m & ( agn ) + 14 & & & & & & & n11l snr + 15 & y & y & y & 0.50 & 045448.37 - 662102.9&sed rises in mid - ir & ( agn ) + 16 & & & & & & & n11l snr + 17 & y & y & y & 0.59 & 045448.95 - 663120.0&sed peaking at 3@xmath48 m & ( agn ) + 18 & & & & & & & n11l snr + 19 & & & & & & & n11l snr + 20 & & & & & & & n11l snr + 21 & y & y & y & 0.31 & 045452.50 - 662644.0&sed flat in ir & ( agn ) + 22 & & & & & & & n11l snr + 24 & y & y & y & 0.18 & 045454.76 - 662703.8&resolved galaxy with a bright nucleus & agn + 25 & y & y & & 0.61 & 045458.47 - 662240.3 & & galactic k5v+wd at 2.8 kpc + 27 & y & y & y & 0.45 & 045501.88 - 663146.8&faint in optical , brighter in ir & ( agn ) + 28 & y & y & y & 0.65 & 045501.86 - 663122.3 & & galactic g5v star at 1.4 kpc + 33 & & y & y & 0.57 & 045510.07 - 663318.3&sed rising in ir & ( agn ) + 36 & y & y & y & 0.33 & 045518.05 - 661735.2&sed rising in ir & ( agn ) + 37 & & & & 0.10 & 045518.66 - 662156.3 + 38 & y & y & y & 0.21 & 045519.00 - 662701.2 & & agn ( see sect .
+ 43 & y & y & & 1.00 & 045524.15 - 662830.5 & & galactic k7v star at 2.2 kpc + 44 & y & y & & 0.56 & 045524.17 - 662535.1&sed and colors do nt match stars & + 45 & y & y & y & 0.49 & 045524.46 - 662223.4 & & agn ( see sect .
+ 47 & & & & 0.69 & 045526.43 - 662225.8 + 48 & y & y & y & 0.26 & 045529.40 - 662312.1 & & massive star in lmc , elst4 + 50 & & y & & 0.36 & 045531.55 - 662639.8&sed flat in ir & ( agn ) + 53 & & y & y & & & faint ir source with high background & + 54 & y & y & & 0.41 & 045533.65 - 662933.1 & & galactic m4v star at 490 pc + 55 & & & & & & faint ir source with high background & ( agn ) + 56 & y & y & y & & & & galactic k0 star hd268670 + 57 & & y & & 0.66 & 045535.47 - 662736.1&sed flat in ir & ( agn ) + 59 & & y & & 0.43 & 045536.77 - 662526.5 + 61 & y & y & & 0.47 & 045537.76 - 662517.3&sed rises in ir & ( agn ) + 63 & & y & y & 0.34 & 045538.97 - 662550.0&sed rises in ir & ( agn ) + 64 & y & y & y & 0.22 & 045539.71 - 662959.4 & & agn ( see sect .
+ 65 & & y & & 0.91 & 045540.15 - 662012.2&sed rises in ir & ( agn ) + 68 & & & & 0.84 & 045542.37 - 662610.1 + 72 & & y & y & 0.47 & 045549.23 - 662654.0&sed rises in ir & ( agn ) + 77 & y & y & y & 0.80 & 045553.93 - 662459.5 & & massive star in lmc , elst92 + 78 & y & y & & 0.51 & 045554.14 - 662818.5 & & galactic k0v star at 4.1 kpc + 86 & y & y & & 0.67 & 045557.28 - 663130.1&sed flat in ir & ( agn ) + 87 & y & y & & 0.09 & 045558.11 - 661856.1 & & galactic k5v star at 1.5 kpc + 89 & y & & & 0.97 & 045601.72 - 661900.9 + 93 & y + 95 & y & y & & 0.36 & 045604.13 - 662126.6 & & galactic m4v star at 330 pc + 97 & y & y & & 0.18 & 045604.57 - 662358.0 & & massive star in lmc , elst68 + 98 & y & y & y & 0.36 & 045604.93 - 662638.5&sed rises in ir & agn ( see sect .
+ 99 & & y & & 0.71 & 045604.98 - 662807.7&sed rises in ir & ( agn ) + 100 & & & & 0.96 & 045605.11 - 662507.8&sed rises in ir , ir emission extended & + 103 & y & y & y & 0.91 & 045607.36 - 663330.8&sed rises in ir & ( agn ) + 105 & & y & & 0.88 & 045608.56 - 662607.8 & & + 106 & y & & & & & & a0 star in lmc + 107 & y & y & & 0.66 & 045610.31 - 663046.0 & & galactic m2v at 820pc + 108 & y & y & & & & & late b dwarf in lmc + 109 & y & y & & & & & galactic k0v+wd at 4kpc + 110 & y & y & y & 0.54 & 045611.02 - 662824.3 & & massive star in lmc , elst33 + 111 & & y & & & & faint , adjacent to bright source + 115 & y & y & & 0.45 & 045619.29 - 662701.9 & & massive star in lmc , elst65 + 116 & & y & y & 0.49 & 045619.49 - 662630.6 + 117 & & y & y & 0.22 & 045620.15 - 662718.2 & & agn ( see sect .
+ 118 & & y & & 0.34 & 045620.35 - 663009.8&sed rises in ir & ( agn ) + 120 & & y & & 0.57 & 045622.41 - 662837.3 + 121 & & y & & 0.72 & 045622.88 - 663106.1 + 123 & & y & y & 0.13 & 045624.58 - 662206.7 & & agn ( see sect .
+ 126 & y & y & y & 0.34 & 045629.65 - 662138.9 & & massive star in lmc , elst51 + 130 & y & y & & 0.43 & 045632.42 - 663012.5&[hkn2006 ] ob candidate & b2iii star in lmc + 134 & y & y & y & 0.48 & 045634.64 - 662826.8 & & massive group in lmc , hd32228 + 136 & y & y & y & 0.46 & 045636.29 - 663123.0&[hkn2006 ] ob candidate & b2iii star in lmc + 139 & y & y & & & & & late b dwarf in lmc + 141 & y & y & y & 0.73 & 045642.44 - 662518.0 & & massive star in lmc , elst31 + 142 & y & y & y & 0.10 & 045643.19 - 662502.7 & & massive star in lmc , pgmw3070 + 145 & y & & & & & & late b dwarf in lmc , pgmw1356 + 146 & y & y & & 0.56 & 045646.80 - 662446.5 & & massive star in lmc , pgmw3120 + 152 & y & y & & & & & massive group in lmc , pgmw3204/9 + 157 & y & y & & 0.71 & 045700.85 - 662357.3 & & massive star in lmc , elst50 + 159 & y & & & & & sed flat in optical , not detected in ir & + 160 & y & y & y & 0.48 & 045704.04 - 662131.5&sed flat from optical to ir & + 164 & y & y & y & & & off - axis source , star in psf & massive group in lmc , n11a + 165 & & & & & & off - axis source , star in psf & star ?
+ cccccccccc 48 & 004 & & oc9.7ib & n & 0.48 & 5.80@xmath385 & @xmath386 & @xmath387 & @xmath3880.12 + 77 & 092 & & o7v & y & 0.91 & 5.17 & @xmath389 & @xmath390 & @xmath3910.17 + 97 & 068 & & o7v & n & 0.10 & 5.06@xmath385 & @xmath392 & @xmath393 & @xmath3940.13 + 110 & 033 & 1005 & b0iiin & n & 0.23 & 5.07@xmath385 & @xmath395 & @xmath396 & @xmath3970.2 + 115 & 065 & 1027 & o6.5v & n & 0.07 & 5.17@xmath385 & @xmath398 & @xmath399 & @xmath4000.14 + 126 & 051 & & o5vn & n & 0.05 & 5.31@xmath385 & @xmath401 & @xmath402 & @xmath4030.12 + 134@xmath404 & & & wr+o & & 0.00 & 6.29 & @xmath405 * & @xmath406 & @xmath4070.08 + 141 & 031 & 3061 & on2iii & n & 0.65 & 5.84@xmath385 & @xmath408 & @xmath409 & @xmath4100.11 + 142 & & 3070 & o6v & & 0.14 & 5.68 & @xmath411 * & @xmath412&@xmath4130.03 + 146 & & 3120 & o5.5 v & c ? & 0.50 & 5.88 & @xmath414 * & @xmath415 & @xmath4160.07 + 152 * * & & 3204/9 & o & & 0.66 & 6.25 & @xmath417 * & @xmath418&@xmath4190.05 + 157 & 050 & 3224 & o4 - 5+o7 & sb2 & 0.91 & 5.66 & @xmath420 & @xmath421&@xmath4220.10 + 164@xmath423 & 028 & 3264 & o6 - 8v & n & 0.89 & 5.65 & @xmath420 * & @xmath424&@xmath4250.10 + & 038 & 3100 & o5iii & n & 0.67 & 5.69@xmath385 & @xmath401 & @xmath426 & @xmath4270.12 + lllllll source & @xmath78 & @xmath79 & norm & @xmath34 ( dof ) & @xmath428 & @xmath429 + & ( @xmath430@xmath7 ) & ( kev ) & ( @xmath431@xmath432 ) & & + + 134 & @xmath433 & @xmath434 & @xmath435 & 0.88 ( 6 ) & 1.99 & 1.99 + + 142 & @xmath436 & @xmath437 & @xmath438 & 1.14 ( 19 ) & 5.39 & 6.32 + + 146 & @xmath439 & @xmath440 & @xmath441 & 1.88 ( 6 ) & 1.93 & 2.84 + + 152 & @xmath442 & @xmath443 & @xmath444 & 0.83 ( 10 ) & 3.40 & 6.31 + + 164 & @xmath445 & @xmath446 & @xmath447 & 1.84 ( 2 ) & 2.16 & 4.27 + + llcccccc 007 & & o8 ib ( f ) & n & 5.72 & 0.58 & @xmath852.02 & @xmath4487.00 + 010 & 1310 & o9.5 iii + b1 - 2 : & sb2 & 5.42 & 0.29 & @xmath853.92 & @xmath4486.41 + 011 & & oc9.5 ii & sb1 & 5.49 & 0.43 & @xmath850.66 & @xmath4487.25 + 013 & 3223 & o8 v & sb2 & 5.66 & 0.48 & @xmath852.43 & @xmath4486.86 + 018 & 3053 & o6 ii(f@xmath449 ) & n & 5.64 & 0.43 & @xmath851.42 & @xmath4487.07 + 019 & & o8 - 9 iii - v((f ) ) & sb2 & 5.40 & 0.24 & @xmath851.68 & @xmath4486.76 + 020 & & o5 i(n)fp & y & 5.50 & 0.14 & @xmath854.32 & @xmath4486.45 + 022 & & o6.5 ii(f ) var & n & 5.45 & 0.29 & @xmath850.98 & @xmath4487.04 + 026 & & o2.5 iii(f@xmath450)&n & 5.92 * & 0.26 & @xmath856.64 & @xmath4486.68 + 029 & & oc9.7 ib & n & 5.21 * & 0.34 & @xmath850.42 & @xmath4487.17 + 032 & 3168 & o7 ii(f ) & n & 5.43 * & 0.38 & @xmath851.69 & @xmath4486.79 + 041 & & o6.5 iaf & y & 5.17 & 0.19 & @xmath852.53 & @xmath4486.35 + 043 & 1519 & o7 iii @xmath451 b0 : & sb2 & 5.12 & 0.14 & @xmath851.69 & @xmath4486.48 + 045 & & o9 - 9.5 iii & n & 5.15 * & 0.26 & @xmath851.97 & @xmath4486.44 + 046 & & o9.5 v & y & 4.91 & 0.05 & @xmath853.94 & @xmath4485.90 + 049 & 1110 & o7.5 v & sb1 & 5.04 & 0.07 & @xmath850.70 & @xmath4486.78 + 052 & & o9.5 v & sb2 & 4.95 & 0.19 & @xmath859.15 & @xmath4485.57 + 058 & & o5.5 v((f ) ) & n & 5.27 * & 0.12 & @xmath852.11 & @xmath4486.53 + 059 & 1125 & o9 v & sb1 & 4.84 & 0.05 & @xmath850.99 & @xmath4486.43 + 060 & 3058 & o3 v((f@xmath450 ) ) & n & 5.57 * & 0.53 & @xmath851.27 & @xmath4487.05 + 061 & & o9 v & n & 5.20 * & 0.50 & @xmath854.08 & @xmath4486.17 + 063 & & o9 : vn & sb2 & 4.88 & 0.22 & @xmath858.87 & @xmath4485.52 + 071 & & o8 : v & sb2 & 4.84 & 0.19 & @xmath854.22 & @xmath4485.80 + 080 & 3173 & o7 : v @xmath451 o9 : & sb2 & 4.91 & 0.29 & @xmath851.27 & @xmath4486.39 + 087 & & o9.5 vn & n & 4.91 * & 0.38 & @xmath850.99 & @xmath4486.50 + 091 & & o9 v & y & 4.78 & 0.41 & @xmath852.82 & @xmath4485.91 + 108 & & o9.5 v & n & 4.56 & 0.19 & @xmath852.39 & @xmath4485.76 + 122 & & o9.5 v & n & 4.37 & 0.00 & @xmath850.84 & @xmath4486.03 + 123 & & o9.5 v & n & 4.58 * & 0.02 & @xmath851.41 & @xmath4486.01 + & 1194 & o9.5:iv & & 4.92 & 0.29 & @xmath850.84 & @xmath4486.58 + & 1200 & o6:v : p & c ? & 5.65 & 0.89 & @xmath851.13 & @xmath4487.18 + & 1239 & o7-o8:v : & & 4.90 & 0.24 & @xmath851.27 & @xmath4486.38 + & 1288 & o9 v & & 4.45 & 0.12 & @xmath851.13 & @xmath4485.98 + & 1292 & o9:iii : & & 4.67 & 0.77 & @xmath851.13 & @xmath4486.20 + & 1363 & o8.5 iaf & & 5.63 & 0.41 & @xmath851.55 & @xmath4487.02 + & 1365 & o9 v & & 4.47 & 0.05 & @xmath851.27 & @xmath4485.95 + & 1377 & o8:v & & 4.73 & 0.12 & @xmath851.41 & @xmath4486.16 + & 1388 & o9.7 iab & & 5.70 & 0.36 & @xmath851.69 & @xmath4487.06 + & 1396 & o7 vp & & 5.24 & 0.14 & @xmath851.55 & @xmath4486.63 + & 1431 & o9 v & & 4.58 & 0.14 & @xmath851.27 & @xmath4486.06 + & 1481 & o7 v((f ) ) & & 5.22 & 0.17 & @xmath851.83 & @xmath4486.54 + & 1483 & o6.5 iii(f ) & & 5.59 & 0.24 & @xmath851.83 & @xmath4486.91 + & 1486 & o6.5 v & & 5.07 & 0.14 & @xmath851.55 & @xmath4486.46 + & 3016 & o9.5:v & & 4.68 & 0.34 & @xmath851.13 & @xmath4486.21 + & 3045 & o9.5 iii & & 4.55 & 0.36 & @xmath851.13 & @xmath4486.08 + & 3073 & o6.5 v & & 5.00 & 0.41 & @xmath851.83 & @xmath4486.32 + & 3089 & o8 v & & 4.91 & 0.36 & @xmath851.13 & @xmath4486.44 + & 3102 & o7 v & & 5.44 & 0.41 & @xmath851.83 & @xmath4486.76 + & 3103 & o9.5:iv : & & 4.29 & 0.29 & @xmath851.41 & @xmath4485.72 + & 3115 & o9 v & & 4.67 & 0.41 & @xmath851.97 & @xmath4485.96 + & 3123 & o8.5 v & & 4.92 & 0.38 & @xmath851.41 & @xmath4486.35 + & 3126 & o6.5 v & & 5.28 & 0.65 & @xmath851.27 & @xmath4486.76 + lcccccc north & 04:55:53.2&@xmath3666:16:39.4 & [email protected] & 0 & 1165 & @xmath452 + middle & 04:56:19.9&@xmath3666:21:46.2&[email protected] & 0 & 17570 & @xmath453 + middle - w & 04:55:54.6&@xmath3666:22:21.8&[email protected] & 58.9&4158 & @xmath454 + middle - se & 04:56:45.7&@xmath3666:22:07.9&[email protected] & 0 & 4799 & @xmath455 + middle - ne & 04:56:22.6&@xmath3666:20:32.4 & circle(63.7 ) & & 2314 & @xmath456 + lh10 & 04:56:44.8&@xmath3666:24:56.2 & [email protected] & 322.9&752 & @xmath457 + east of lh9 & 04:57:07.8&@xmath3666:26:52.2 & [email protected] & 323.7&2066 & @xmath458 + west of lh9 & 04:56:02.8&@xmath3666:30:28.9&[email protected] & 0 & 5970 & @xmath459 + lh9 * & 04:56:35.6&@xmath3666:28:34.8&[email protected] & 21.3&13985 & @xmath460 + inside lh9 & 04:56:56.3&@xmath3666:28:53.6 & circle(29.5 ) & & 739 & @xmath461 + near hd32228 * & 04:56:34.7&@xmath3666:28:27.2 & annulus(5,15 ) & & 321 & @xmath462 + bkgd 1 & 04:55:29.9&@xmath3666:18:19.7 & circle(66.2 ) & & 1612 & @xmath463 + bkgd 2 & 04:55:23.3&@xmath3666:29:38.6 & circle(60.7 ) & & 1556 & @xmath464 + bkgd 3 & 04:55:48.8&@xmath3666:25:11.6 & circle(54.1 ) & & 1087 & @xmath465 + annulus a & 04:56:56.3&@xmath3666:28:53.6 & annulus(30,55 ) & & 1400 & @xmath466 + annulus b & 04:56:34.7&@xmath3666:28:27.2 & annulus(30,60 ) & & 2363 & @xmath467 + lcccc north & 1 & 1.95 & 1.95 & 0.14 + middle & 1 + 3 & 15.4 & 295.3 & 3.70 + middle - w & 1 + 3 & 4.15 & 62.4 & 3.55 + middle - se & 1 + 3 & 4.76 & 73.6 & 3.70 + middle - ne & 1 + 3 & 1.89 & 52.7 & 5.16 + lh10 & 3 & 0.98 & 2.36 & 1.08 + east of lh9 & 2 + 3 & 1.77 & 26.0 & 2.70 + west of lh9 & 2 + 3 & 7.62 & 154 . & 7.74 + lh9 & 2 + 3 & 24.3 & 24.4 & 0.53 + inside lh9 & a & 0.26 & 0.26 & 0.12 + near hd32228 & b & 0.23 & 1.35 & 2.71 + | a very sensitive x - ray investigation of the giant region n11 in the lmc was performed using the _ chandra x - ray observatory_. the 300ks observation reveals x - ray sources with luminosities down to 10@xmath0ergs@xmath1 , increasing by more than a factor of 5 the number of known point sources in the field . amongst these detections
are 13 massive stars ( 3 compact groups of massive stars , 9 o - stars and one early b - star ) with @xmath2 to @xmath3 , which may suggest that they are highly magnetic or colliding wind systems . on the other hand , the stacked signal for regions corresponding to undetected o - stars yields @xmath4 , i.e. , an emission level comparable to similar galactic stars despite the lower metallicity .
other point sources coincide with 11 foreground stars , 6 late - b / a stars in n11 , and many background objects .
this observation also uncovers the extent and detailed spatial properties of the soft , diffuse emission regions but the presence of some hotter plasma in their spectra suggests contamination by the unresolved stellar population . |
the purpose of this paper is to address the computation and analysis of one - loop shifts to the masses of semi - local planar topological solitons arising in a natural generalization of the abelian higgs model . seen in ( 3 + 1)-dimensional space - time these solitons become semi - local strings whereas their masses give the string tensions , see @xcite-@xcite . at the critical point that marks the phase transition between type i and type ii superconductivity ,
the set of semi - local self - dual topological solitons is an interesting @xmath3-dimensional moduli space , where @xmath4 is the number of quanta of the magnetic flux , see @xcite-@xcite .
recently , superconducting semilocal ( non self - dual ) strings with very intriguing properties have been discovered in this model @xcite .
computations of one - loop mass corrections will be performed using the heat kernel / zeta function regularization method .
the high - temperature asymptotic expansion of the heat function , see @xcite-@xcite-@xcite-@xcite , is a powerful tool that was applied for the first time to the calculation of kink mass shifts in @xcite -the @xmath5 susy case- and @xcite -the non susy case- .
one - loop corrections to @xmath6 supersymmetric self - dual nielsen - olesen vortices were computed in a similar approach by vassilevich and rebhan - van nieuwenhuizen - wimmer in references @xcite and @xcite . in the second paper
, the authors also showed that the central charge of the susy algebra is modified in one - loop order in such a way that the bogomolny bound is saturated at the semi - classical level . sometime later , we calculated the one - loop mass shift for @xmath7 ( non - susy ) self - dual no vortices carrying a quantum of magnetic flux in reference @xcite .
mass shifts have been given for spherically symmetric self - dual vortices ( when several solitons of a quantum of flux have coinciding centers ) in @xcite up to four magnetic flux quanta .
cruder approximations were also provided for the mass shift of two separated self - dual no vortices -each of them with a quantum of magnetic flux- as a function of the inter - center distance . in reference
@xcite , we studied the one - loop correction to the energy of a degenerate manifold of kinks that arise in a very interesting family of models with two real scalar fields in ( 1 + 1)-dimensions .
these field theoretical systems are obtained through dimensional reduction -plus a reality condition- of an @xmath5 supersymmetric wess - zumino model with two chiral super - fields , see @xcite-@xcite-@xcite .
a comparison between the mass shifts of these composite kinks and the correction to the mass of the ordinary @xmath8 kink was offered in @xcite . in this paper
, we address a similar , but more difficult , situation in ( 2 + 1)-dimensions , comparing one - loop mass corrections of the topological solitons that arise in two planar abelian gauge systems ; one with two complex scalar fields and the other with a single complex scalar field .
the methodology used to accomplish this task is explained in detail in reference @xcite , where a complete list of references can be found .
our paper is organized as follows : in section
2 we describe the model and develop perturbation theory around one of the vacua .
a one - loop renormalization is also performed .
section 3 is devoted to summarizing the structure of the moduli space of self - dual topological solitons . as a novelty
, we also apply a variation of the de vega - shaposnik method @xcite to find numerical solutions for spherically symmetric topological solitons . in section
4 , we explain how to obtain one - loop mass shifts in terms of generalized zeta functions of the second - order differential operators ruling the small fluctuations of the bosonic and ghost fields , and in 5 the high - temperature expansion of the heat kernel is used to give the final formula for one - loop mass shifts of semi - local self - dual topological solitons after application of mellin s transforms .
we present our results in section 6 by means of mathematica calculations of the coefficients of the asymptotic series giving the heat functions .
finally , in the appendix we offer several tables where these coefficients are shown .
the semi - local abelian higgs model @xcite describes the minimal coupling between an @xmath1-gauge field @xmath9 and a doublet @xmath10 of complex scalar fields in a phase where the higgs mechanism takes place .
the term semilocal refers to the fact that while the global symmetry of this system is @xmath11 only the @xmath1 factor is gauged . defining non - dimensional space - time variables , @xmath12 , and fields , @xmath13 , @xmath14 ,
from the vacuum expectation value of the higgs field @xmath15 and the @xmath1-gauge coupling constant , @xmath16 , the action for the semilocal abelian higgs model in ( 2 + 1)-dimensions reads : @xmath17 \qquad , \ ] ] where the covariant derivative is defined as @xmath18 .
the physical spectrum differs from that of the standard ahm model in that , along with the massive vector boson and higgs scalar , there is a complex goldstone field .
the parameter @xmath19 measures the ratio between the square of the masses of the higgs , @xmath20 , and vector particles , @xmath21 .
here , @xmath22 is the higgs field self - coupling .
we choose a system of units where @xmath23 , but @xmath24 has dimensions of length @xmath25 mass . in order to obtain the feynman rules , we expand the action around a classical vacuum .
we shall profit from the @xmath0 global invariance and we choose a vacuum @xmath26 with real upper and vanishing lower components .
the shift of the fields @xmath27 corresponds to the identification of @xmath28 , @xmath29 , and @xmath30 , respectively as higgs , real , and complex goldstone fields .
the choice of the feynman-t hooft r - gauge @xmath31 requires a faddeev - popov determinant to restore unitarity , which amounts to introducing a complex ghost field @xmath32 .
all this together allows us to write the action in the form @xmath33a_\nu + \partial_\mu\chi^*\partial^\mu \chi- \chi^*\chi\right.\\ & + & \frac{1}{2}\partial_\mu g\partial^\mu g-\frac{1}{2 } g^2 + \frac{1}{2}\partial_\mu h\partial^\mu h-\frac{\kappa^2}{2 } h^2+\partial_\mu\varphi^ * \partial^\mu \varphi\\ & - & { \kappa^2\over 2}h ( h^2+g^2)+ a_\mu ( \partial^\mu h g-\partial^\mu g h)+h ( a_\mu a^\mu -\chi^ *
\chi ) + i a_\mu(\varphi^ * \partial^\mu \varphi-\varphi \partial^\mu \varphi^*)\\&+ & \left .
|\varphi|^2 -\frac{\kappa^2}{8 } ( h^2+g^2)^2 + \frac{1}{2}(g^2+h^2 ) a_\mu a^\mu -\frac{\kappa^2}{2}|\varphi|^2(|\varphi|^2+h^2+g^2 + 2h)\right]\end{aligned}\ ] ] in order to deduce the feynman rules for the propagators and vertices shown in tables 1 and 2 .
the propagator of the complex goldstone boson plus two trivalent and four tetravalent vertices accounting for the interactions of goldstone - anti - goldstone pairs must be added to the feynman rules of the abelian higgs model in the feynman-t hooft gauge .
nevertheless , in order to guarantee that all physical quantities will be ultraviolet finite there is still a remaining piece to be added : the action for the counter - terms .
we shall compute this up to one - loop order in the sequel .
lccc + _ particle _ & _ field _ & _ propagator _ & _ diagram _ + + higgs & @xmath34 & @xmath35 & + real goldstone & @xmath36 & @xmath37 & + complex goldstone & @xmath38 & @xmath39 & + ghost & @xmath40 & @xmath37 & + vector boson & @xmath41 & @xmath42 & + llllllll + _ vertex _ & _ weight _ & _ vertex _ & _ weight _ & _ vertex _ & _ weight _ & _ vertex _ & _ weight _
+ + & @xmath43 & & @xmath44 & & @xmath45 & & @xmath46 + & @xmath47 & & @xmath48 & & @xmath49 & & @xmath50 + & @xmath51 & & @xmath47 & & @xmath52 & & @xmath47 + & @xmath53 & & @xmath54 & & @xmath55 & & @xmath55 + + all ultraviolet divergences come from the integral : @xmath56 we also note that even if there are massless particles propagating in 2 + 1 dimensions , @xmath57 is infrared convergent .
the renormalization of the semilocal ahm requires the computation of the following one - loop graphs : 1 .
higgs boson tadpole : + @xmath58 @xmath58 @xmath58 @xmath58 @xmath59 + @xmath60 2 .
higgs boson self - energy : + @xmath58 @xmath58 @xmath58 @xmath58 @xmath59 + @xmath61 3 . real goldstone boson self - energy : + @xmath58 @xmath58 @xmath58 @xmath58 @xmath59 + @xmath61 4 .
complex goldstone boson self - energy : + @xmath58 @xmath58 @xmath58 @xmath58 @xmath59 + @xmath62 5 .
vector boson self - energy : the potentially divergent part is + @xmath58 @xmath58 @xmath58 @xmath58 + @xmath63g^{\mu\nu } + \mbox{finite part } \qquad .\ ] ] in ( 2 + 1)-dimensions the graphs above are the only ultraviolet divergent diagrams for any number of loops in the diagrams , not only in one - loop order , and the theory is super - renormalizable .
thus , in a minimal subtraction scheme we get rid off all ultraviolet divergences arising in the vacuum sector of the model by adding the counter - terms @xmath64\cdot\ , \left[\phi^*_1(x^\mu)\phi_1(x^\mu)+\phi^*_2(x^\mu)\phi_2(x^\mu)-1 \right]\\{\cal l}_{c.t.}^a & = & -\hbar [ i(1)+i(0)]\cdot\ , a_\mu(x^\mu ) a^\mu(x^\mu)\end{aligned}\ ] ] .counter - term vertices [ cols="<,<,^",options="header " , ]
this work has been partially financed by the spanish ministerio de educacion y ciencia and the junta de castilla y leon under grants : fis2006 - 09417 , and va013c05 . | a formula is derived that allows the computation of one - loop mass shifts for self - dual semilocal topological solitons . these extended objects , which in three spatial dimensions are called semi - local strings , arise in a generalized abelian higgs model with a doublet of complex higgs fields . having a mixture of global , @xmath0 , and local ( gauge ) , @xmath1 , symmetries , this weird system may seem bizarre , but it is in fact the bosonic sector of electro - weak theory when the weak mixing angle is @xmath2 . the procedure for computing the semi - classical mass shifts
is based on canonical quantization and heat kernel / zeta function regularization methods .
pacs : 03.70.+k;11.15.kc;11.15.ex _ keywords _ : semi - local topological solitons ; heat - kernel / zeta function regularization ; one - loop shifts to soliton masses |
experiments on a variety of novel materials , ranging from quasi - one - dimensional ( 1d ) mx solids @xcite , organics @xcite and quasi-2d high-@xmath0 cuprates @xcite to 3d colossal - magnetoresistive manganites @xcite , provide clear evidence for the existence of polaronic carriers , i.e. , quasiparticles consisting of an electron and a surrounding lattice distortion .
this has motivated considerable theoretical efforts to achieve a better understanding of strongly coupled electron - phonon systems in the framework of microscopic models .
unfortunately , even for highly simplified models , such as the spinless holstein model @xcite considered here , no exact analytical solutions exist , except for the holstein polaron problem with a relativistic dispersion @xcite or in infinite dimensions @xcite . as a consequence ,
numerous numerical studies have been carried out , focussing either on the empty band limit ( i.e. , one or two electrons only ; see @xcite and references therein ) , or on the half - filled band case in one dimension , where the peierls transition takes place @xcite .
in contrast , very little work has been done at finite carrier densities away from half filling @xcite , which are , however , often realized in experiment @xcite .
recently , this so - called _ many - polaron problem _ has been addressed numerically @xcite .
the results have led to a fairly good understanding of many aspects , but their interpretation is not always straight forward , which makes analytical calculations along these lines highly desirable . in this paper , we propose an analytical approach to the 1d spinless holstein model , capable of describing finite charge - carrier concentrations at arbitrary model parameters , including the important adiabatic intermediate - coupling ( ic ) regime . first ,
using standard perturbation theory based on the self - energy calculation , the spectral functions of charge carriers will be determined in the weak - coupling ( wc ) and the strong - coupling ( sc ) limits at zero temperature ( @xmath1 ) . in the sc regime ,
the relation between the spectral function of polarons , determining the equilibrium properties ( in particular , the chemical potential ) , and the electronic spectral function , determining the photoemission spectrum , will be discussed .
special emphasis will be laid on the distinction between the coherent and the incoherent parts of the spectra , which may be calculated separately within the present approach .
furthermore , using a generalization of the lang - firsov canonical transformation method @xcite , an interpolation scheme between the extreme wc and sc cases will be proposed . in particular ,
the canonical transformation will depend on the distance @xmath2 characterizing the charge distribution across the polaron volume .
for a given set of model parameters and carrier concentration @xmath3 , @xmath2 will be determined from the minimum of the total energy given by the transformed hamiltonian in the first , hartree - like approximation . with @xmath2 found in this way , the polaronic and electronic spectral functions will be calculated to study their dependence on the model parameters and the carrier density .
the results will be discussed with regard to recent numerical calculations @xcite , which have revealed a cross - over from a system with polaronic carriers to a rather metallic system with increasing band filling @xmath3 in the intermediate electron - phonon coupling regime .
the paper is organized as follows . in section [ sec : model ]
we introduce the model used , whereas in section [ sec : gfapproach ] we derive the analytical results for wc , sc and ic regimes .
our results are discussed in section [ sec : results ] , and section [ sec : conclusion ] contains our conclusions .
in this paper , we are exclusively concerned with the holstein model ( hm ) of spinless fermions , which describes electrons coupled locally to einstein phonons .
although we shall use different canonical transformations to describe the ic and sc regime later on , the hamiltonian can be written in the general form @xmath4 where the definition of @xmath5 and @xmath6 will be different depending on the approach used . here
@xmath7 ( @xmath8 ) creates ( annihilates ) a spinless fermion at site @xmath9 , @xmath10 and @xmath11 are bosonic operators for the dispersionless phonons of energy @xmath12 ( @xmath13 ) , and the strength of the electron - phonon interaction is specified by the dimensionless coupling constants @xmath14 and @xmath15 in the adiabatic ( @xmath16 ) and anti - adiabatic ( @xmath17 ) regimes , respectively , where @xmath18 is the well - known polaron binding energy in the atomic limit [ @xmath19 for @xmath20 in equation ( [ eq : hm ] ) ] . in the wc case , in which we use the original , untransformed holstein hamiltonian
, we have @xmath21 , where @xmath22 denotes the chemical potential , and non - zero coefficients @xmath23 in contrast , the starting point in the sc regime will be the hamiltonian with @xmath24 and @xmath25 we treat the hm ( [ eq : hm ] ) using the formalism of the generalized matsubara green functions introduced by kadanoff and baym @xcite and bonch - bruevich and tyablikov @xcite , and applied to the single - polaron problem by schnakenberg @xcite .
the green function equation of motion deduced from @xmath26 may be converted into an equation for the self - energy of the spinless fermions , and can be solved by iteration ( see @xcite for details ) . in the second iteration step ,
the self - energy is obtained in the form @xmath27
\,,\end{aligned}\ ] ] where @xmath28 represents the first - order fermionic matsubara green function , and the symbol @xmath29 denotes the time ordering operator acting on the imaginary times @xmath30 .
fourier - transforming both sides of equation ( [ eq : sea ] ) , carrying out the standard summation over the matsubara boson frequencies @xcite , and using the analytical continuation in the complex frequency plane , the retarded momentum and energy - dependent fermion green function follows as @xmath31 where @xmath32 is the fermionic band dispersion in the first approximation , @xmath33 ( @xmath34 ) , and @xmath35 denotes the fourier transform of the collisional part of the self - energy given by the second term on the r. h. s of equation ( [ eq : sea ] ) .
the related , normalized spectral function is given by @xmath36 although the insulating peierls phase with long - range charge - density - wave order , which is the ground state of the half - filled spinless holstein model above a critical coupling strength depending on @xmath12 @xcite , can in principle be incorporated into the present theory , this has not been done here , as we are interested in intermediate band fillings away from @xmath37 .
furthermore , we neglect any extended pairing for @xmath38 , as well as the possible formation of a polaronic superlattice in the half - filled band case ( @xmath37 ) . finally , the possibility of phase separation , which can in principle occur in the present model , is not considered . in the wc limit
the self - energy is given by @xmath39 \,,\ ] ] with the fermi function @xmath40 and the bare bandwidth @xmath41 . at @xmath1 , and defining @xmath42 , we get @xmath43\end{aligned}\ ] ] and @xmath44 \right.\nonumber\\ + \left .
\int_{\xi_0}^w \frac{\rmd\xi}{\sqrt{1-(\xi / w)^2 } } \delta[{\omega}- { \omega}_0 - ( \xi-\mu ) ] \right ] \,.\end{aligned}\ ] ] in the sequel , we shall distinguish between coherent and incoherent contributions to the single - particle spectral functions , as defined by a zero and non - zero imaginary part of the self - energy , respectively .
the coherent part of the spectrum is given by @xmath45 \,,\ ] ] where the renormalized band energy is the solution of @xmath46 \,,\end{aligned}\ ] ] with @xmath47 ( the bare band dispersion ) , and the spectral weight takes the form @xmath48
\bigg| \,.\end{aligned}\ ] ] for the incoherent part of the spectral function we find @xmath49^{\frac{1}{2}}\int_{-w}^{\xi_0}\rmd\xi\delta({\omega}+{\omega}_0+\mu-\xi ) } { \left[w^2-({\omega}+{\omega}_0+\mu)^2\right]\left({\omega}+\mu-\xi_k-\mathrm{re}\,\sigma(k,{\omega})\right)^2 + ( { \omega}_0{e_\mathrm{p}})^2}\ ] ] and @xmath50^{\frac{1}{2}}\int_{\xi_0}^w\rmd\xi\delta({\omega}-{\omega}_0+\mu-\xi ) } { \left[w^2-({\omega}-{\omega}_0+\mu)^2\right]\left({\omega}+\mu-\xi_k-\mathrm{re}\,\sigma(k,{\omega})\right)^2 + ( { \omega}_0{e_\mathrm{p}})^2 } \,.\ ] ] finally , the chemical potential @xmath22 for a given electron density @xmath3 is determined by @xmath51 hamiltonian ( [ eq : hm ] ) with the coefficients ( [ eq : scc ] ) represents the hamiltonian of small polarons , which are the correct quasiparticles in the sc limit . using the procedure outlined above
, we obtain the polaron self - energy as @xmath52\,.\end{aligned}\ ] ] at @xmath1 , we have @xmath53 , @xmath54\end{aligned}\ ] ] and @xmath55 \right . \nonumber\\ \left .
+ \theta(-{\omega}-s{\omega}_0)\delta[{\omega}+s{\omega}_0-(\xi+\eta ) ] \right\ } \,.\end{aligned}\ ] ] here @xmath56 is the heaviside step function and we have used the definition @xmath57 + \frac{(g^2)^s}{s!}\cos(2k ) \,.\ ] ] the coherent part of the spectrum , non - zero for @xmath58 , is given by @xmath59\ ] ] with the renormalized band energy @xmath60 being the solution of @xmath61 where @xmath62 , and the spectral weight takes the form @xmath63
\bigg| \,.\end{aligned}\ ] ] the imaginary part of the self - energy , determining the incoherent excitations , is non - zero only for @xmath64 .
consequently , we get @xmath65 with @xmath66 if @xmath67 , for a given @xmath68 , only one term of the sum in equation ( [ eq : im_si_sc ] ) contributes . the corresponding index @xmath69 is determined by the conditions @xmath70 or @xmath71 the incoherent spectrum then consists of non - overlapping parts : @xmath72 } \,.\end{aligned}\ ] ] here we have defined @xmath73 the polaron spectral function determines the equilibrium properties of the spinless hm in the sc regime .
the photoemission spectra , however , are determined by the electron spectral function which is related to the retarded green function containing electronic operators .
according to the canonical lang - firsov transformation @xcite , which defines small polaron states , the relation between the polaronic operators @xmath74 entering the hamiltonian ( [ eq : hm ] ) with the coefficients ( [ eq : scc ] ) , and the transformed electron operators @xmath75 , reads @xmath76 c^{{\phantom{\dag}}}_j \,,\quad { \widetilde{c}}^\dag_j = \exp[-g(b^\dag_j - b^{{\phantom{\dag}}}_j ) ] c^\dag_j\ ] ] or , in the bloch representation ,
@xmath77 c^{{\phantom{\dag}}}_j \,,\quad { \widetilde{c}}^\dag_k = \frac{1}{\sqrt{n}}\sum_j \rme^{\rmi k r_j } \exp[-g(b^\dag_j - b^{{\phantom{\dag}}}_j ) ] c^\dag_j \,.\ ] ] we start our derivation of the relation between the polaronic and electronic spectra from the time - ordered green function @xcite for the ( transformed ) electron operators @xmath78 and factorize the statistical averages with respect to polaron and phonon variables @xmath79 } \rme^{-g[b^\dag_{j+d}(0)-b^{{\phantom{\dag}}}_{j+d}(0 ) ] } { \rangle}\,,\ ] ] where @xmath80 and @xmath81 represents the time - ordered green function of polaron operators fulfilling @xmath82 the averages over the phonon variables in equation ( [ gtk ] ) will be evaluated using mutually independent local einstein oscillators having the time - dependence @xmath83 .
working in the low - temperature approximation we obtain @xmath84\end{aligned}\ ] ] with @xmath85 and @xmath86 , and the convergence factor @xmath87 , @xmath34 . introducing the generalized function @xmath88^{-1}$ ] ,
@xmath34 , we get @xmath89\,.\end{aligned}\ ] ] the green functions @xmath90 are related to the polaron spectral function @xmath91 through @xmath92 { a_\mathrm{p}}(k,{\omega } ) \,.\ ] ] at @xmath1 , we have @xmath93 \,.\end{aligned}\ ] ] the relation between the time - ordered green function @xmath94 and the associated retarded green function @xmath95 in the low - temperature approximation reads @xmath96 and hence we obtain @xmath97 of course , equations ( [ rgrgk ] ) and ( [ ragt ] ) hold for the electron green function as well .
consequently , the electron spectral function @xmath98 is expressed in terms of the polaron spectral function @xmath91 as @xmath99 \,.\end{aligned}\ ] ] similar results have been derived before in @xcite . as we shall see in section [ sec : results ] , the results of the wc ( sc ) approximation are in good agreement with numerical calculations @xcite if @xmath100 ( @xmath101 ) .
however , the cross - over between these limiting cases , revealed by the numerical calculations , appears to be out of reach for the analytical formulae hitherto deduced . to interpolate between wc and sc , we shall modify the method of canonical transformation by lang and firsov @xcite . in the latter ,
the term of the hm ( [ eq : hm ] ) linear in the local oscillator coordinate @xmath102 is completely eliminated by the translational transformation @xmath103 $ ] . as a result
, the local lattice oscillator at the site @xmath9 is shifted by @xmath104 , if occupied by a charge carrier , whereas there is no such deformation at unoccupied sites . to generalize this picture ,
we abandon the site localization of both the charge carrier and the lattice deformation in the transformation .
physically , these localizations will be destroyed with increasing hopping rate and charge - carrier concentration in the ic regime .
different canonical transformations , taking into account charge - density - wave order at @xmath37 , have been proposed , e.g. , by zheng @xcite .
however , in this approach , there is no filling dependence of their variational parameter and of the mean lattice deformation , which is crucial for a correct description of the adiabatic ic regime . the probability for the charge carrier to be found at a distance @xmath105 from the center @xmath106 of the polaron
will be assumed to be proportional to @xmath107 . in one dimension , and setting the lattice constant to unity , we use the normalized distribution @xmath108 accordingly , the shift of the local oscillator with coordinate @xmath109 is assumed to be @xmath110 with @xmath111 the last term in equation ( [ eq : shift ] ) , characterizing the mean lattice deformation background , takes into account the influence of nearest - neighbor sites only .
the canonical transformation leading to the oscillator shift ( [ eq : shift ] ) reads @xmath112 \,,\quad { \overline{\gamma}}= k - \gamma \,.\ ] ] carrying out the transformation @xmath113 for the hamiltonian ( [ eq : hm ] ) with ( [ eq : wcc ] ) , the terms of @xmath114 containing polaron operators are modified with the following coefficients [ cf .
equations ( [ eq : wcc ] ) and ( [ eq : scc ] ) ] @xmath115 moreover , we have @xmath116 $ ] . owing to numerical problems which occur in certain parameter regimes when using the full hamiltonian with the coefficients ( [ eq : trans_ic ] ) , the variational parameter @xmath2 of the transformation will be determined in the first approximation , which is analogous to the hartree approximation .
the corresponding polaron spectral function is given as @xmath117 $ ] , where @xmath118 with @xmath119 $ ] , and @xmath5 as defined above .
@xmath2 is then defined by the position of the minimum of the total energy per site @xmath120 in the first approximation , i.e. , @xmath121 with the condition for @xmath22 @xmath122 the last term in equation ( [ eq : energ ] ) arises in @xmath114 from the lattice deformation background at finite concentration @xmath3 .
@xmath2 determined in this way for each set of model parameters will be used to calculate both the electron and polaron spectral function , taking into account the multi - phonon processes included in @xmath114 [ cf .
equation ( [ eq : trans_ic ] ) ] . using the same procedure as in the sc case ,
we calculate the self - energy and spectral function at @xmath1 , finding @xmath123 \nonumber\\ \lo- 2{\overline{\gamma}}(1-{\overline{\gamma}})\frac{{e_\mathrm{p}}}{\pi } \mathcal{p}\int_{-{\widetilde{w}}}^{{\widetilde{w}}}\frac{\rmd\xi}{\sqrt{1-(\xi/{\widetilde{w}})^2 } } \left(\frac{\xi}{{\widetilde{w}}}+\cos{k}\right ) \nonumber\\ \times \left [ \frac{\theta(\xi+\eta)}{{\omega}-{\omega}_0-(\xi+\eta ) } - \frac{\theta(-\xi-\eta)}{{\omega}+{\omega}_0-(\xi+\eta ) } \right ] \nonumber\\ \lo+ ( 1-{\overline{\gamma}})^2\frac{{e_\mathrm{p}}\omega_0}{\pi{\widetilde{w } } } \mathcal{p}\int_{-{\widetilde{w}}}^{{\widetilde{w}}}\frac{\rmd\xi}{\sqrt{1-(\xi/{\widetilde{w}})^2 } } \nonumber\\ \times \left [ \frac{\theta(\xi+\eta)}{{\omega}-{\omega}_0-(\xi+\eta ) } + \frac{\theta(-\xi-\eta)}{{\omega}+{\omega}_0-(\xi+\eta ) } \right]\end{aligned}\ ] ] and @xmath124 here @xmath125 and @xmath126 , @xmath127 are defined as in equations ( [ eq : f_sc ] ) and ( [ eq : fpm ] ) but with @xmath128 replaced by @xmath129 .
note that the above equations reproduce the corresponding sc results in the limit @xmath130 ( @xmath131 ) , and also the wc ones in the limit @xmath132 ( @xmath133 ) .
the coherent spectrum is again given by equations ( [ eq : sc_coherent ] ) and ( [ eq : sc_band ] ) , with the spectral weight @xmath134_{\omega = e_k+\eta } \right| \,,\ ] ] whereas the incoherent part takes the familiar form @xmath135 ^ 2 + [ \mathrm{im}\sigma(k,\omega)]^2 } \,.\ ] ] finally , equation ( [ eq : sc_el_spectrum ] ) , which determines the relation between the electronic and polaronic spectral functions , also applies to the present case if @xmath128 is replaced by @xmath129 throughout .
( colour online ) coherent ( @xmath136 , @xmath137 ) and incoherent ( @xmath138 , @xmath139 ) parts of the spectral function in the weak - coupling approximation for different band fillings @xmath3 . here
@xmath140 and @xmath141.,title="fig:",scaledwidth=49.5% ] ( colour online ) coherent ( @xmath136 , @xmath137 ) and incoherent ( @xmath138 , @xmath139 ) parts of the spectral function in the weak - coupling approximation for different band fillings @xmath3 . here
@xmath140 and @xmath141.,title="fig:",scaledwidth=49.5% ] + ( colour online ) coherent ( @xmath136 , @xmath137 ) and incoherent ( @xmath138 , @xmath139 ) parts of the spectral function in the weak - coupling approximation for different band fillings @xmath3 . here
@xmath140 and @xmath141.,title="fig:",scaledwidth=49.5% ] ( colour online ) coherent ( @xmath136 , @xmath137 ) and incoherent ( @xmath138 , @xmath139 ) parts of the spectral function in the weak - coupling approximation for different band fillings @xmath3 . here
@xmath140 and @xmath141.,title="fig:",scaledwidth=49.5% ] as in section [ sec : gfapproach ] , we first discuss the wc and sc limits , before turning to the important ic regime . since we use a finite number of momenta @xmath142 ,
it is not possible to tune the band filling @xmath3 ( via the chemical potential @xmath22 ) to a specific , desired value with arbitrary accuracy . in order to simplify the discussion of the different density regimes , we therefore report rounded values of @xmath3 in the figures and the text .
the largest deviations of the actual @xmath3 from the value reported occur in the sc case for which , however , the density dependence is very weak ( see section [ sec : res : strong - coupling ] ) . figures [ fig : wc_density_w0.4][fig : coupling_wc ] show the electronic spectral function @xmath143 obtained from the wc approximation .
the coherent spectrum ( @xmath144 ) is given as the solution of equations ( [ eq : wc_a_coh])([eq : wc_zk ] ) with energies @xmath145 .
for @xmath146 , @xmath138 , calculated according to equations ( [ eq : wc_a_inc1 ] ) and ( [ eq : wc_a_inc2 ] ) , consists of peaks having widths proportional to @xmath18 . a comparison of figures [ fig : wc_density_w0.4 ] ( a ) and [ fig : wc_density_w2.0](a ) shows the spreading of the coherent spectrum with increasing @xmath12 .
finally , comparing figure [ fig : wc_density_w0.4](d ) [ [ fig : wc_density_w2.0](b ) ] with figure [ fig : coupling_wc](a ) [ [ fig : coupling_wc](b ) ] , we observe a broadening of the peaks in @xmath138 with increasing @xmath18 .
the evolution of the spectral function with increasing carrier density @xmath3 is illustrated in figure [ fig : wc_density_w0.4](a)(d ) .
the coherent part is shifted inside the spectrum as a function of @xmath3 ( i.e. , with @xmath22 ) . additionally , the shape of @xmath138 is also affected by @xmath3 , due to the dependence of equations ( [ eq : wc_a_inc1 ] ) and ( [ eq : wc_a_inc2 ] ) on the chemical potential @xmath22 . in figure
[ fig : weight_wc ] , we plot the total spectral weight @xmath147 , as well as the coherent weight @xmath148 .
note that for any @xmath149 , the coherent band @xmath60 is restricted to the interval @xmath150 $ ] , so that the corresponding coherent weight is significantly smaller than the @xmath151 value of unity .
this reduction is less pronounced for larger phonon frequency @xmath152 ( figure [ fig : weight_wc ] ) .
furthermore , we see that with increasing @xmath18 , the sum rule for @xmath143 becomes more and more violated , as expected for a wc approximation ( for a more detailed discussion of sum rules see section [ sec : res : strong - coupling ] ) .
finally , figure [ fig : ek_wc ] displays the coherent band dispersion @xmath60 at small carrier density @xmath153 . as known from the single - polaron problem , the coherent weight @xmath154 drops to zero as the bare phonon dispersion intersects with the renormalized band .
this gives rise to a flattening of the coherent band at large @xmath142 @xcite , which is well reproduced by the simple wc approximation .
( colour online ) as in figure [ fig : wc_density_w0.4 ] , but for @xmath155 .
, title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : wc_density_w0.4 ] , but for @xmath155 . ,
title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : wc_density_w0.4 ] , but for @xmath156 and fixed band filling @xmath157 . here
( a ) @xmath140 and ( b ) @xmath155 . ,
title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : wc_density_w0.4 ] , but for @xmath156 and fixed band filling @xmath157 . here
( a ) @xmath140 and ( b ) @xmath155 . , title="fig:",scaledwidth=49.5% ] coherent ( @xmath137 ) and total ( @xmath139 ) spectral weight from @xmath143 ( see text ) for @xmath140 ( @xmath158 , @xmath159 ) and @xmath155 ( @xmath160 , @xmath161 ) as a function of the polaron binding energy @xmath18 . here
@xmath162 . , scaledwidth=49.5% ] renormalized band @xmath60 for @xmath140 , @xmath141 and @xmath153 .
symbols indicate non - zero coherent spectral weight @xmath154 .
the horizontal line ( @xmath137 ) corresponds to @xmath163.,scaledwidth=49.5% ] ( colour online ) coherent ( @xmath164 , @xmath137 ) and incoherent ( @xmath165 , @xmath139 ) parts of the polaron spectral function in the strong - coupling approximation . here
@xmath157 , @xmath166 and ( a ) @xmath140 , ( b ) @xmath155 . , title="fig:",scaledwidth=49.5% ] ( colour online ) coherent ( @xmath164 , @xmath137 ) and incoherent ( @xmath165 , @xmath139 ) parts of the polaron spectral function in the strong - coupling approximation . here
@xmath157 , @xmath166 and ( a ) @xmath140 , ( b ) @xmath155 . , title="fig:",scaledwidth=49.5% ] renormalized band @xmath60 for @xmath157 , @xmath140 and different values @xmath167 ( @xmath160 ) and @xmath166 ( @xmath158).,scaledwidth=49.5% ] coherent ( polaronic , @xmath137 ) and total ( @xmath139 ) spectral weight from @xmath91 for @xmath140 ( @xmath158 , @xmath159 ) and @xmath155 ( @xmath160 , @xmath161 ) , as a function of the polaron binding energy @xmath18 . here
@xmath162 . , scaledwidth=49.5% ] ( colour online ) as in figure [ fig : polaron_spectrum_sc ] , but showing the electronic spectral functions @xmath168 ( @xmath137 ) and @xmath169 ( @xmath139).,title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : polaron_spectrum_sc ] , but showing the electronic spectral functions @xmath168 ( @xmath137 ) and @xmath169 ( @xmath139).,title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : coupling_sc ] , but for different band fillings @xmath3 . here
@xmath140 and @xmath167.,title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : coupling_sc ] , but for different band fillings @xmath3 . here
@xmath140 and @xmath167.,title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : sc_density_w0.4 ] , but for @xmath155.,title="fig:",scaledwidth=49.5% ] ( colour online ) as in figure [ fig : sc_density_w0.4 ] , but for @xmath155.,title="fig:",scaledwidth=49.5% ] we now turn to the opposite , sc limit . the theory presented in section [ sec : strong - coupling ] directly yields the polaronic spectrum @xmath91 , results for which are shown in figure [ fig : polaron_spectrum_sc](a ) for @xmath140 and @xmath166 .
the spectrum is dominated by a coherent polaronic band with negligible width ( for the dependence of the bandwidth on @xmath18 see figure [ fig : ek_sc ] ) having a spectral weight @xmath154 close to unity ( cf .
figure [ fig : weight_sc ] ) .
this suggests that small polarons are the correct quasiparticles in the sc regime .
note that opposite to the wc case , where the sum rule for the spectral function becomes more and more violated with increasing coupling ( figure [ fig : weight_wc ] ) , here the sc approximation becomes increasingly better with increasing @xmath18 ( figure [ fig : weight_sc ] ) .
we would like to point out that such changes in the total spectral weight are absent in the work of alexandrov and ranninger @xcite , since the latter was restricted to the lowest ( first ) order of the self - energy , similar to the hartree approximation discussed in section [ sec : interm - coupl ] .
in general , the total spectral weight contained in the ( electronic or polaronic ) spectral function for given parameters depends on the approximations made . as illustrated by figures [ fig : weight_wc ] and [ fig : weight_sc ] , the total spectral weight approaches the exact value of unity in the wc and sc regimes , respectively , so that the normalization of the spectrum serves as a measure of the validity of the underlying approximations . since in the present case even the first moment ( i.e. , the normalization ) shows deviations from exact results , we have refrained from checking the more complicated sum rules derived in @xcite .
this is also true of the ic case discussed below .
the effect of increasing the phonon frequency @xmath12 can be seen by comparing figures [ fig : polaron_spectrum_sc](a ) and ( b ) . most noticeably , for larger @xmath12 , the width of the coherent polaron band roughly scaling proportional to @xmath170is larger .
the polaronic spectrum is related to the electronic spectrum @xmath171 by equation ( [ eq : sc_el_spectrum ] ) , and typical results in the adiabatic and non - adiabatic regimes are shown in figure [ fig : coupling_sc ] .
although strictly speaking a distinction between coherent and incoherent contributions can not be made in the case of @xmath171 [ cf .
equation ( [ eq : sc_el_spectrum ] ) ] , it is useful to separate the two terms on the r. h. s. of equation ( [ eq : sc_el_spectrum ] ) , and to identify the first as the contribution of the coherent polaron band . carrying out the transformation from @xmath172 to @xmath171 according to equation ( [ eq : sc_el_spectrum ] ) , the weight of the coherent polaron band visible in figure [ fig : polaron_spectrum_sc ] approximately acquires a prefactor @xmath170 .
the remaining contributions to the electronic spectrum correspond to phonon - assisted photoemission processes . in the case of @xmath171 ,
the main difference between the adiabatic [ figure [ fig : coupling_sc](a ) ] and the non - adiabatic regime [ figure [ fig : coupling_sc](b ) ] is the significantly larger weight of the coherent band for large @xmath12 since @xmath173 .
consequently , the weight contained in the incoherent excitations is noticeably reduced .
the results in figures [ fig : sc_density_w0.4 ] and [ fig : sc_density_w2.0 ] show a certain dependence on the band filling @xmath3 , but there occur no qualitative changes even at ic @xmath167 .
this is in contrast to recent numerical work @xcite . in particular
, the spectrum in figure [ fig : sc_density_w0.4](b ) is substantially different from figures [ fig : ic_w0.4](c ) and ( d ) , which are all for the same parameters .
we shall see below that a more satisfactory description of the real physics can be obtained using the variational approach discussed in section [ sec : interm - coupl ] .
total energy per site @xmath120 as a function of the variational parameter @xmath2 for different values of @xmath18 from the hartree approximation .
here @xmath157 and ( a ) @xmath140 , ( b ) @xmath155 .
the values of @xmath174 are ( from top ) ( a ) 0.1 , 1 , 2 , 2.5 , 3 , 4 , ( b ) 0.1 , 1 , 2 , 4 , 6 , 8 , and 12 .
, title="fig:",scaledwidth=30.0% ] total energy per site @xmath120 as a function of the variational parameter @xmath2 for different values of @xmath18 from the hartree approximation .
here @xmath157 and ( a ) @xmath140 , ( b ) @xmath155 .
the values of @xmath174 are ( from top ) ( a ) 0.1 , 1 , 2 , 2.5 , 3 , 4 , ( b ) 0.1 , 1 , 2 , 4 , 6 , 8 , and 12 .
, title="fig:",scaledwidth=30.0% ] ( a ) total energy per site @xmath120 as a function of @xmath18 obtained from @xmath143 ( see text ) , where @xmath157 and @xmath140 . also shown
are quantum monte carlo results at inverse temperature @xmath175 obtained with the method of @xcite .
( b ) exact diagonalization ( ed ) and density matrix renormalization group ( dmrg ) results for the renormalized kinetic energy as a function of carrier density @xmath3 for @xmath167 , @xmath140 and different cluster sizes @xmath176 .
, title="fig:",scaledwidth=30.0% ] ( a ) total energy per site @xmath120 as a function of @xmath18 obtained from @xmath143 ( see text ) , where @xmath157 and @xmath140 . also shown
are quantum monte carlo results at inverse temperature @xmath175 obtained with the method of @xcite .
( b ) exact diagonalization ( ed ) and density matrix renormalization group ( dmrg ) results for the renormalized kinetic energy as a function of carrier density @xmath3 for @xmath167 , @xmath140 and different cluster sizes @xmath176 .
, title="fig:",scaledwidth=30.0% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath140 .
the parameters are ( a ) @xmath141 , @xmath177 , ( b ) @xmath166 , @xmath178 and ( c ) @xmath167 , @xmath179 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](a ) .
panel ( d ) shows cluster perturbation theory results for photoemission [ @xmath180 , dashed lines ] and inverse photoemission [ @xmath181 for @xmath157 and @xmath182 ( taken from @xcite).,title="fig:",scaledwidth=49.5% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath140 .
the parameters are ( a ) @xmath141 , @xmath177 , ( b ) @xmath166 , @xmath178 and ( c ) @xmath167 , @xmath179 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](a ) .
panel ( d ) shows cluster perturbation theory results for photoemission [ @xmath180 , dashed lines ] and inverse photoemission [ @xmath181 for @xmath157 and @xmath182 ( taken from @xcite).,title="fig:",scaledwidth=49.5% ] + ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath140 .
the parameters are ( a ) @xmath141 , @xmath177 , ( b ) @xmath166 , @xmath178 and ( c ) @xmath167 , @xmath179 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](a ) .
panel ( d ) shows cluster perturbation theory results for photoemission [ @xmath180 , dashed lines ] and inverse photoemission [ @xmath181 for @xmath157 and @xmath182 ( taken from @xcite).,title="fig:",scaledwidth=49.5% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath140 .
the parameters are ( a ) @xmath141 , @xmath177 , ( b ) @xmath166 , @xmath178 and ( c ) @xmath167 , @xmath179 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](a ) .
panel ( d ) shows cluster perturbation theory results for photoemission [ @xmath180 , dashed lines ] and inverse photoemission [ @xmath181 for @xmath157 and @xmath182 ( taken from @xcite).,title="fig:",scaledwidth=49.5% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath155 .
the parameters are ( a ) @xmath141 , @xmath183 , ( b ) @xmath184 , @xmath185 , ( c ) and ( d ) @xmath166 , @xmath186 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](b ) . in the white regions in ( d ) @xmath187 ( see text ) .
, title="fig:",scaledwidth=49.5% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath155 .
the parameters are ( a ) @xmath141 , @xmath183 , ( b ) @xmath184 , @xmath185 , ( c ) and ( d ) @xmath166 , @xmath186 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](b ) . in the white regions in ( d ) @xmath187 ( see text ) .
, title="fig:",scaledwidth=49.5% ] + ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath155 .
the parameters are ( a ) @xmath141 , @xmath183 , ( b ) @xmath184 , @xmath185 , ( c ) and ( d ) @xmath166 , @xmath186 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](b ) . in the white regions in ( d ) @xmath187 ( see text ) .
, title="fig:",scaledwidth=49.5% ] ( colour online ) spectral functions from the variational approach for different values of @xmath18 , @xmath157 and @xmath155 .
the parameters are ( a ) @xmath141 , @xmath183 , ( b ) @xmath184 , @xmath185 , ( c ) and ( d ) @xmath166 , @xmath186 .
the values of @xmath2 have been chosen according to figure [ fig : free_energy_ic](b ) . in the white regions in ( d ) @xmath187 ( see text ) .
, title="fig:",scaledwidth=49.5% ] as seen in the preceding sections , both the wc and sc approximations are not capable of accounting for the recently discussed carrier density - driven cross - over from a polaronic system to a metallic system with phonon - dressed electrons @xcite .
in fact , the electronic spectrum always remains wc / sc - like in character . in order to obtain a reasonable , analytical description of the ic regime
, we therefore use the variational approach proposed in section [ sec : interm - coupl ] . figure [ fig : free_energy_ic ] shows results for the total energy per site as a function of the variational parameter @xmath2 for different values of @xmath18 .
the evolution is very similar to the large - to - small polaron cross - over in the one - electron case @xcite .
note that the present results have been obtained using the hartree approximation , as discussed in section [ sec : interm - coupl ] . for weak coupling @xmath18 , in the adiabatic case [ figure [ fig : free_energy_ic](a ) ]
, we find a minimum in the total energy at a finite value of @xmath2 . upon
increasing @xmath18 , a second local minimum starts to develop near @xmath130 , associated with the small - polaron state which becomes the ground state in the sc limit .
the jump of the optimal value of @xmath2 from a large to a small value at a critical @xmath18 suggests a first - order transition from an extended to a small - polaron state .
however , such a sharp transition is absent in the single - polaron case , and not expected for @xmath188 either .
the discontinuous cross - over appearing in our results for @xmath189 is a consequence of the approximation used .
in contrast , in the non - adiabatic regime [ figure [ fig : free_energy_ic](b ) ] , there exists only a single minimum , which shifts to smaller @xmath2 with increasing coupling @xmath18 .
moreover , compared to figure [ fig : free_energy_ic](a ) , the dependence of the total energy on @xmath2 is much weaker .
apart from the comparison of the spectral function with other methods presented below , variational approaches often not capable of yielding dynamic properties are usually judged by the total energy as opposed to exact data .
figure [ fig : e0_ep](a ) shows the second - order results for the total energy as a function of @xmath18 , using the optimal values of the parameter @xmath2 as determined from the hartree approximation .
clearly , the agreement between our ic approach and results from quantum monte carlo @xcite is very good at weak and strong coupling ( the variational approach reproduces the wc and sc limits of sections [ sec : weak - coupling ] and [ sec : strong - coupling ] ) , whereas there are notable deviations at ic .
the missing higher - order corrections causing the violation of the sum rule discussed earlier
lead to generally overestimated values of the energy .
note that the agreement of the energy with exact results is even better in the non - adiabatic regime ( not shown ) .
to illustrate the density - driven cross - over from ( large ) polarons to slightly dressed electrons ( scattered by diffusive phonons ) , figure [ fig : e0_ep](b ) reports exact numerical results for the renormalized kinetic energy .
the latter may serve as a measure for the carrier mobility .
the increase as a function of @xmath3 may be interpreted as originating from the overlap of the displacement clouds surrounding the carriers and , finally , the dissociation of the polaronic quasiparticles at large @xmath3 .
we again begin the discussion of the spectral functions with the adiabatic case @xmath140 .
following the discussion of section [ sec : interm - coupl ] , we determine the optimal @xmath2 from the position of the minimum of the total energy for given @xmath18 and @xmath3 ( figure [ fig : free_energy_ic ] ) .
figures [ fig : ic_w0.4](a ) and ( b ) resemble closely to the wc and sc results of figures [ fig : wc_density_w0.4](d ) and [ fig : sc_density_w0.4](b ) , respectively .
however , for ic @xmath167 [ figure [ fig : ic_w0.4](c ) ] , we find a rather metallic spectrum with a broad main band crossing the fermi level , and with low - energy excitations available .
the corresponding polaronic spectrum ( not shown ) reveals that the variational approach correctly predicts the absence of well - defined polaronic quasiparticles , as suggested by the non - negligible incoherent contributions in @xmath91 , lying close in energy to the coherent band .
this is in contrast to the sc case , where polaronic quasiparticles dominate , and in which the coherent band is well separated from the incoherent excitations .
note that in figure [ fig : ic_w0.4](c ) [ and also in figures [ fig : ic_w2.0](c ) and ( d ) ] the incoherent part becomes slightly negative for @xmath190 ( @xmath191 ) at large ( small ) @xmath142 , which is an artifact of our approximation .
remarkably , the overall features of the spectrum in figure [ fig : ic_w0.4](c ) are very similar to the corresponding numerical results in @xcite , reproduced in figure [ fig : ic_w0.4](d ) . finally , in the non - adiabatic regime , it has been found numerically that small polarons remain the correct quasiparticles even at large band fillings @xmath3 @xcite .
again , the variational approach is able to describe the physics correctly . in particular , figures [ fig : ic_w2.0](c ) and ( d )
reveal that a dominant coherent band , well separated from incoherent excitations and having a relatively small bandwidth , exists even at ic . as for the wc and sc cases discussed above ,
we have also checked the total spectral weight in the present case .
we find that the deviations from the exact value of unity are largest for wc , whereas the sum rule is fulfilled to within 1 - 10% ( depending on @xmath12 and @xmath18 ) at ic and sc .
we have presented an analytical treatment of the one - dimensional spinless holstein model based on calculations of the self - energy in the framework of the generalized matsubara functions . to connect the analytical results to previous numerical ones ,
the electronic spectral function determining the photoemission spectrum has been computed for finite carrier concentrations in dependence on the electron - phonon coupling strength and the phonon frequency . in the strong - coupling limit , the electron spectral function has been deduced from the spectral function of small polarons .
however , it was shown that the electron picture and the small - polaron picture both cease to be correct if we approach the intermediate - coupling regime from the weak and strong - coupling side , respectively . to describe the cross - over from the strong to the weak - coupling limit ,
an interpolation scheme based on a generalization of the lang - firsov canonical transformation has been proposed .
the latter was defined for each set of model parameters by a distance @xmath2 which characterizes the charge distribution across the polaron volume .
the parameter @xmath2 , as deduced from the minimum of the total energy in the first approximation , was shown to increase with decreasing @xmath18corresponding to a cross - over from large to small polarons and , at the same time , the coupling dependence of @xmath2 was found to be stronger in the adiabatic than in the non - adiabatic case .
the spectral functions calculated at weak , strong and intermediate coupling are in a good agreement with recent numerical calculations . moreover , analytical results deduced by means of the self - energy calculations enabled us to distinguish between the coherent and incoherent parts of the spectrum .
most importantly , starting from the strong - coupling limit , it was shown that the spectral weight of the incoherent polaron spectrum increases with decreasing coupling @xmath18 , and that the energy separation of the incoherent peaks from the coherent spectrum is continuously reduced in the adiabatic regime .
on the contrary , the coherent part of the electronic ( photoemission ) spectra is reduced by a factor @xmath170 , and the incoherent part representing the phonon - assisted photoemission processes becomes increasingly dominant with increasing coupling .
finally , at intermediate coupling and finite carrier densities , our results support recent numerical findings which suggest that the system can no longer be described in terms of ( small ) polaronic quasiparticles @xcite .
we want to thank f. x. bronold , h. sormann and g. wellein for helpful discussions , and a. s. alexandrov for valuable comments on the manuscript .
furthermore , we would like to thank g. hager for providing us with the dmrg data shown in figure [ fig : e0_ep](b ) .
this work was supported by the austrian science fund project p15834 , the dfg through spp1073 , the dfg and the academy of sciences of the czech republic ( ascr ) under grant nr .
436 tse 113/33/0 - 2 , and the bavarian konwihr .
one of us ( mh ) is grateful to hpc - europa .
mh and hf acknowledge the hospitality of the institute of physics , ascr , prague . | an analytical approach to the one - dimensional spinless holstein model is proposed , which is valid at finite charge - carrier concentrations .
spectral functions of charge carriers are computed on the basis of self - energy calculations .
a generalization of the lang - firsov canonical transformation method is shown to provide an interpolation scheme between the extreme weak and strong - coupling cases .
the transformation depends on a variationally determined parameter that characterizes the charge distribution across the polaron volume .
the relation between the spectral functions of polarons and electrons , the latter corresponding to the photoemission spectrum , is derived .
particular attention is paid to the distinction between the coherent and incoherent parts of the spectra , and their evolution as a function of band filling and model parameters .
results are discussed and compared with recent numerical calculations for the many - polaron problem . |
half the interstellar medium in late type spiral galaxies is in molecular gas clouds , and about 90% by number of the atoms within these clouds are tied up in the h@xmath0 molecule .
stars form within molecular clouds , and the pure rotational lines of h@xmath0 are important coolants enabling cloud collapse .
the cooling radiation from these lines is also important to the formation of the giant molecular clouds ( gmcs ) themselves . despite its abundance and importance ,
very little is known about cool h@xmath0 on galactic scales .
this is because h@xmath0 has no dipole moment , so that dipole radiation from its low - lying rotational energy levels is forbidden .
the low lying levels instead radiatively decay by emitting relatively weak quadrupole ( @xmath5j = 2 ) radiation .
therefore the pure rotational lines of h@xmath0 are challenging to observe .
three of the low lying lines : the s(1 ) ( 17.0@xmath6 m ) , s(2 ) ( 12.3@xmath6 m ) , and the s(4 ) ( 8.0@xmath6 m ) lines are transmitted through the earth s atmosphere , and have been detected from regions as diverse as supernovae remnants @xcite , photo - dissociation regions ( pdrs ) associated with ob star formation sites ( e.g. * ? ? ?
* ) , and disks enveloping young stellar objects ( e.g. * ? ? ?
. however , detection of the lowest lying s(0 ) ( 28.2@xmath6 m ) line , and detection of these lines from cool molecular clouds in external galaxies awaited the advent of the short wavelength spectrometers ( sws ) on the infrared space observatory ( iso , e.g. * ? ? ? * ) .
a few years later , the infrared spectrograph ( irs ) of the spitzer space telescope became available and the detection of these lines from a wide variety of external galaxies became routine ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
h@xmath0 exists in two varieties : para ( nuclear spins anti - aligned ) and ortho ( spins aligned ) . the relative ortho - to - para ( @xmath7 )
abundances reflect the local gas temperature since proton exchange reactions with h@xmath8 , h and h@xmath9 are effective in thermalizing the @xmath7 ratio ( * ? ? ? * and references therein ) . at high temperatures
( t@xmath10200 k ) the o / p ratio is set by the ratio of statistical weights at 3:1 . since there is no radiative coupling between the ortho and para species , when analyzing radiative transitions they can be treated as independent molecules . the lowest lying transitions for the para species are : j=2 - 0 ( s(0 ) , 28.2@xmath6 m ) , and 4 - 2 ( s(2 ) , 12.3@xmath6 m ) , and for the ortho species : j=3 - 1 ( s(1 ) , 17.0@xmath6 m ) , and j=5 - 3 ( s(3 ) , 9.7@xmath6 m ) . with their very small einstein a coefficients the h@xmath0 level populations
are easily thermalized at molecular cloud densities , and the emitted lines are nearly always optically thin .
therefore , the line intensities trace the column density of the emitting level , and since the line emitting levels are thermalized and lie hundreds of degrees above the ground state ( larger than typical molecular gas temperatures ) , the line ratios within a species should reflect the gas temperature , t@xmath11 .
for example , the s(2)/s(0 ) line ratio gets larger by a factor of 150 as t@xmath11 goes from 70 to 100 k. the para j=2 level lies just 510 k above the ground so it can be emitted by cool ( t@xmath12 k ) molecular gas and is an important coolant for molecular clouds in , for example , the early stages of collapse into protostars ( t@xmath4100 k , n@xmath13100@xmath14 ) .
are the rotational lines of h@xmath0 detectable from normal star forming galaxies , and if so , is the emitting gas important ?
work with the iso - sws demonstrated that the answer to both questions is an emphatic yes . large scale s(0 ) and s(1 ) line emission is reported both from the face - on sc galaxy ngc 6946 @xcite , and the edge on sb galaxy ngc891 @xcite .
the strong emission indicates surprisingly large masses of warm molecular gas .
spitzer irs studies confirm that strong h@xmath0 rotational line emission and large masses of warm molecular gas are common in a wide variety of galaxies including nuclei of starburst galaxies @xcite and ulirgs @xcite , in disk of galaxies @xcite , extragalactic regions such as stephan s quintet @xcite , as well as tidal dwarf galaxies @xcite .
is a nearby ( d = 9.5 mpc @xcite ) sb galaxy that is presented to us nearly edge - on and is thought to be a close milky way analogue . since an edge - on galaxy presents the largest column densities to our telescopes , and the h@xmath0 rotational lines are optically thin , ngc891 provides the best opportunity to study warm h@xmath0 in the outer reaches of galaxies . with the iso - sws
, @xcite detected the s(0 ) and s(1 ) lines at eight positions along the plane of ngc891 .
the lines were surprisingly strong even in the outer regions of the stellar disk .
assuming @xmath7 = 3 they find warm ( t@xmath15130 to 230 k ) molecular cloud component pervades the galactic disk with a total mass comparable to both the atomic hi and the cold molecular gas as traced through its co emission : m@xmath16 @xmath17m@xmath18 @xmath4 m@xmath19 .
it would be a surprise if half the molecular ism in the galaxy is warm , and it is a challenge to provide a heating mechanism for this gas .
furthermore , based on a drop in the s(1)/s(0 ) line intensity ratio at radii greater than 8 kpc , @xcite argue for an additional cool ( t@xmath20 90 k ) component which becomes dominant at large radii . with @xmath7 @xmath42 to 3 , the cooler component is massive : m@xmath21 @xmath4 5 - 15 @xmath22 m@xmath19 in the outer disk .
the cool molecular gas mass is so large that it can be a substantial fraction of the `` dark matter '' required to drive the rotation curve out to the edge of the opitical disk of ngc891 .
this is particularly intriguing due to theoretical work that suggests a large fraction of the `` dark '' baryonic matter in galaxies may be in co free h@xmath0 clouds ( e.g. * ? ? ? * ) .
the detection of a massive cooler component , if proven correct , is a very significant result .
these results need verification so we have repeated the observations of the s(0 ) and s(1 ) lines along the plane of ngc891 using the more sensitive spectrograph on the spitzer space telescope @xcite . in addition , the important s(2 ) line is included in the large instantaneous bandwidth of the spitzer irs .
the addition of the s(2 ) line mapping provides a key element in tracing the gas excitation .
the galaxy was observed on 2004 august 7 as part of the irsguaranteed time ( gt ) program with the two high resolution ( r@xmath4600 ) modules , short - high ( sh ) and long - high ( lh ) .
we also obtained supplemental , deep integrations in selected positions through a spitzer cycle 4 open time ( ot ) program with pid 40877 .
for all observations , the sh slits covers the 9.7 - 19.5@xmath6 m range and it is 4.7@xmath23 in size , while lh covers the 18.8 - 37.2@xmath6 m range and its size is 11.1@xmath24 .
the width of each slit was designed so that it is equal to the first airy ring of an unresolved point source at their short wavelength ( see * ? ? ? * ) .
we obtained spectra for a total of 27 positions symmetrically sampled around the nucleus of the galaxy at a uniform spacing of 30@xmath25 ( 1.53 kpc ) with a position angle of 23 degree along the galactic plane of ngc891 .
six positions off the galactic plane to the northwest were also included to search for possible extended h@xmath26 emission .
for each pointing , the on - source integration time was 60 and 28 secs for sh and lh respectively .
in addition two off positions to the northeast and northwest away from the galaxy , which were used for background subtraction , were also obtained with twice the integration time .
the irs pointings are indicated in figure 1 with circles of 25@xmath27 in diameter , slightly larger than the length of the lh slit .
the enumeration of the pointings follows the notation of table 1 , which provides the ra and dec of each pointing along with its distance in arcsec along the disk of the galaxy from the central pointing ( id=1 ) . within the ot program
, we obtained on - source integration times of 750 and 900 seconds for 9 positions ( positions 1,6,8,10,12,20,22,24 and 26 in the notation of figure 1 ) and 5 positions ( positions 1,6,8,20,22 ) in sh and lh respectively .
we also re - observed the off possitions for background subtraction .
given the fact that irs has no moving parts and that the long axis of the sh slit is perpendicular to that of the lh slit , we could not modify at will the orientation of the irs slits on the sky .
the slit orientation is determined by the ecliptic coordinates of the science target and the time of observations . by properly selecting the time of the observations
, we were able to ensure that for the sh observations , which contain the s(1 ) and s(2 ) lines , the sh slit was placed with the long axis perpendicular to the plane of the galaxy .
the lh observations , which trace the s(0 ) line , were obtained at the same time , with the lh slit placed with the long axis along the plane of the galaxy , since the two slits are perpendicular .
the data were processed by the spitzer science center ( ssc ) pipeline version 17.2 .
as it is discussed in detail in the irs manual , for each pointing , two observations are made by placing the target on two nod positions along the slit .
the 128@xmath22128 pixel detectors are read with @xmath28 successive non destructive reads before the final n - th read which also resets the array .
for our observations n=16 for sh and 8 for lh .
the ssc pipeline starts with the 128@xmath29 sh data cubes and converts them to a 128@xmath30 two - dimensional image after linearization correction , subtraction of darks , cosmic - ray removal , stray light and flat field correction .
each image , called basic calibrated data ( bcd ) contains the ten echelle orders of the sh and lh slits respectively .
our analysis started from these bcd files . at each pointing , we subtracted the dedicated off - source background image in order to remove the sky emission from our slit .
then , the two dimensional images were cleaned with the irsclean package to remove the bad pixels as well as to correct the so called rogue " pixels .
these are pixels whose values depend not only on the photons they collect , but also on the total number of photons hitting the whole 128@xmath30 array , in a nonlinear manner finally , the cleaned images were used to extract the spectra using the full slit extraction method of the irs data analysis package smart @xcite .
the outputs from smart produced one spectrum at each nod position , and thene the two nod position spectra are combined .
the data from the edge pixels within each echelle order were manually cleaned by truncating a few pixels at the beginning and end of each order and when there was inconsistency , the blue end of the order was preferred .
no scaling was needed between the adjacent grating orders within the same module .
the final errors include both statistical and systematic effects , and we report them in table 1 .
the overall absolute spectrophotometric uncertainty of irs is 5% , so that spectrophotometric uncertainty is small compared with the signal - to - noise ratio of our spectra .
table 1 lists the observed line fluxes and upper limits at all positions , while the locations on the galaxy where these spectra were obtained superposed on an 8@xmath6 m spitzer / irac image of the galaxy , are indicated in figure 1 .
we also plot in figure 2 the spectra around the s(0 ) , s(1 ) , and ( s3 ) h@xmath26 lines from locations where at least one line has been detected .
both the s(0 ) and s(1 ) lines are clearly detected at 17 of the positions along the galactic plane .
figure 3a plots the line intensity distributions along the plane of galaxy .
the lines have fairly uniform intensities in the inner galaxy , with some enhancement at the nucleus , beginning to fall off at 6.1 kpc ne and 4.6 kpc sw of the nucleus to values below our detection threshold at regions beyond 13.8 kpc ne and sw of the nucleus .
the s(2 ) line was only detected in 11 positions , all within 10 kpc of the nucleus , but this is very likely a sensitivity issue .
the detection limit is similar for all three lines , but the s(2 ) is typically 4 times fainter than the s(0 ) line and 1.7 times fainter than the s(1 ) line .
therefore , its distribution could well be nearly identical to the lower j lines .
the line intensity ratios are also fairly constant across the galaxy , with the exception of enhanced s(1 ) line emission at the nucleus .
we do not detect h@xmath0 line emission off the plane of the galaxy .
however , our integration time is rather modest , so perhaps this is not too surprising : @xcite obtained substantially deeper integrations ( nearly 6.5 hours , or 400 times longer than ours ) at two positions 1 kpc off the disk , where they clearly detected the s(1 ) line .
based on these results @xcite estimate the scale height of the s(1 ) line emission is @xmath48.4 to 9.9 " ( 390 - 490 pc , based on a distance of 9.5 mpc ) . correcting their results for the expected extinction of the lines by dust , they estimate that the intrinsic width of the warm molecular gas is @xmath44.8 to 5.7@xmath25 ( 220 to 260 pc ) .
it is important to compare our spitzer irs results to those from the iso sws study reported in @xcite .
howevever both the iso and spitzer maps are sparsely sampled , and neither was obtained with a large enough beam to detect all of the extra - planar emission if it is as extensive as the results of @xcite suggest .
we first correct for the missing extra - planar emission , by assuming the apparent ( exponential ) scale height for the h@xmath0 line emission is the average of @xcite s values : 9.1@xmath31 .
we took spitzer lh ( s(0 ) line ) data with the slit width extending @xmath325.55@xmath31 perpendicular to the plane , and sh ( s1 ) and s(2 ) line data with the long axis of the slit expending @xmath325.65@xmath31 perpendicular to the plane of the galaxy .
for these cases , we calculate that 45% and 46% of the line flux ( integrated perpendicular to the plane ) is detected within our lh and sh beams respectively .
the ( rectangular ) iso beams were 20@xmath33 and 14@xmath34 for the s(0 ) and s(1 ) lines respectively , and these beams were tilted so that the long axis of the rectangle traced the plane of ngc 891 @xcite . for these cases , and again comparing against an assumed scale height of 9.1@xmath31 and integrated perpendicular to the plane , we calculate that 45% and 53% of the line flux was detected by the iso sws spectrometer .
we apply these extra - planar corrections to both data sets for proper comparisons .
we also must correct for flux not detected along the plane of the galaxy , since neither ours , nor the iso data sets are fully sampled .
the iso sws slit along the plane of the galaxy ( point - to - point , along the diagonal ) was 34@xmath31 and 24@xmath31 for the s(0 ) and s(1 ) lines respectively . since
the spitzer data are only sampled every 30@xmath27 , we must interpolate between spizer samplings to properly compare with the iso data . for both the s(0 ) and s(1 ) data sets , we interpolate linearly between the spitzer sampled regions , then integrate within a region covering the angular extent of the iso beam along the plane .
this should be a good estimator for most of the galaxy , especially for the s(0 ) line , where the iso and spitzer beam sizes are nearly the same .
a possible exception is the nucleus in the s(1 ) line , where the relatively much smaller spitzer beam may emphasize any strong central peak .
figure 3b illustrates the comparison between spitzer and iso observations including all the effects described above . for the inner ( r @xmath35 7 kpc ) galaxy our spitzer line fluxes are about a factor 1.55 and 1.85 larger than those of iso in the s(0 ) and s(1 ) lines respectively , but the spitzer s(1)/s(0 ) line ratio is not much different than the iso s(1)/s(0 ) ratio .
therefore , we derive similar excitation temperatures for the inner galaxy as those presented in @xcite(see section 4.1.1 below ) . however , in the outer galaxy ( r @xmath36 7 kpc ) the s(1)/s(0 ) line ratios diverge .
the spitzer s(1 ) line flux does not fall off nearly as rapidly as that measured by iso , so that the spitzer / iso ratio grows to @xmath4 2.57 in the outer regions .
in addition , the spitzer s(0 ) line flux falls off more rapidly in the outer galaxy than does the iso s(0 ) line flux , so that spitzer / iso ratio gets smaller , to @xmath4 0.81 in the outer regions .
therefore , unlike our spitzer results , where the s(1)/s(0 ) line ratio is nearly constant over the disk , the iso results show a marked decrease in the s(1)/s(0 ) ratio for the 3 regions sampled at r @xmath36 7 kpc from the nucleus .
it is this change in line ratios for the outer regions that led @xcite to invoke a second , cool gas contributor to the observed s(0 ) line emission .
beam size is not the source of this discrepancy : since the iso beam has its diagonal along the plane of the galaxy , it was equally sensitive to extraplaner emission as the spitzer beam .
furthermore , it is not just the added s(0 ) line flux in the outer disk , but it is also the dearth of s(1 ) line radiation in the outer disk reported in the iso observations that changes the observed s(1)/s(0 ) line ratio , hence the derived gas excitation .
the nearly constant line ratios revealed by the spitzer observations are consistent with warm molecular gas , with near uniform excitation conditions over the galaxy - there is no evidence for an additional cool gas component .
furthermore , as we will show below , this warm molecular gas component is far less massive ( @xmath3513% ) than the cold h@xmath26 emitting molecular component ( @xmath37 m@xmath2 ) invoked by @xcite . to begin an analysis of mid - ir line emission from the edge - on galaxy ngc891 , we need to first make corrections for the non - negligible extinction in the lines due to dust along the line of sight .
the h@xmath0 line emission is likely to be highly correlated with that of the co(1 - 0 ) line which is tightly confined to the plane of the galaxy , so that to make a good extinction correction , we need high spatial resolution proxies .
fortunately , there exist high resolution images of ngc891 in the two primary tracers of the neutral ism the hi 21 cm line which traces atomic hydrogen , and the co(1 - 0 ) line which traces cold molecular hydrogen .
we calculate the total column density in the molecular gas and atomic gas components within our spitzer beams as a function of distance along the plane of the galaxy as follows . for the co(1 - 0 ) distribution
, we use the data contained in figure 2 of @xcite .
these data were obtained with a 2.3@xmath27 beam , but are plotted in their figure 2 by integrating the emission perpendicular to the plane of the galaxy , then refering it to a 1@xmath27 strip along the plane . to compare with our data ,
we therefore re - expand their strip to follow an exponential fall - off perpendicular to the plane of the galaxy with the local co(1 - 0 ) scale height ( also given in figure 2 of @xcite ) .
the observed scale height ( given as full - width - at - half - maximum ) varies and corresponds to a 1/e scale height ranging from about 2.9@xmath27 at the nucleus to @xmath4 4.3@xmath27 in the outer regions . for regions just exterior ( further removed from the nucleus along the galactic plane ) to those explored by @xcite ( r@xmath10200@xmath27 )
we have added in the iram 30 m telescope ( 20@xmath27 beam ) observations of @xcite , scaled assuming the line emission follows the 4.3@xmath27 scale height observed in the outer galaxy by @xcite .
we used a co(1 - 0 ) integrated intensity to h@xmath0 column density conversion factor of x = @xmath38 @xmath39/(k km s@xmath40 ) as for the solar neighborhood @xcite @xmath39/(k km s@xmath40 ) , then the extinction corrected luminosities of the h@xmath0 lines go up 10% , and the overall calculated warm ( h@xmath0 line emitting ) to cold ( co(1 - 0 ) line emitting ) molecular mass fraction goes down to 11% from 16% ( see section 4.2.2 ) .
the derived physical parameters of the gas , and its excitation mechanisms do not change significantly . ] , and derived an expected hydrogenic column density within 1@xmath27 strips parallel to the galactic plane with an exponential fall - off as given in @xcite .
the hi column density is taken from the 21 cm line map of @xcite with a 7@xmath27 beam .
since the scale height of the hi emission is relatively large ( @xmath4 6@xmath27 , @xcite , we estimate the hi column density within our spizer beams by direct inspection of the hi maps . note
that since the hi column is a small fraction of the h@xmath0 column ( as traced by co ) in the the regions of the galaxy where extinction is important ( within @xmath4 5 kpc of the nucleus , see figure 4a ) , a more detailed model of the hi distribution is not warranted .
the total neutral gas mass is nearly equally divided into molecular ( @xmath41 m@xmath2 , scoville et al .
1993 ) , using x = @xmath38 @xmath39/(k km s@xmath40 ) , and atomic ( @xmath42 m@xmath2 * ? ? ?
* ) components .
we plot the calculated column density of hydrogen nuclei derived from the co and hi observations , smoothed to our s(0 ) beam size along the plane of the galaxy in figure 4a .
the column plotted only includes mass within our spitzer beam which extends @xmath32 5.55@xmath27 perpendicular to the galactic plane : the mass missed at high galactic latitude is not included in this plot . the total column density of hydrogen nuclei within our beam peaks at @xmath43 @xmath39 at the nucleus and is above @xmath44 @xmath39 from -10 to + 10 kpc along the galactic plane .
such large columns lead to sizable extinction even for the s(0 ) line at 28@xmath6 m . to find the extinction ,
we take the column density of hydrogen nuclie as traced by the co distribution ( modeled as an exponential fall - off ) and the observed hi distribution within our beam to calculate the expected extinction as a function of distance from the galactic plane .
the extinction calculation is broken up into 1@xmath27 strips parallel to the plane .
for each strip , we first find a@xmath45 , and then we calculate the local extinction correction , c(@xmath46 ) as defined below .
we then average these extinction corrections within our beam to get an estimate for the total correction that must be applied to the observed h@xmath26 line emission .
extinction is estimated using the `` galactic '' dust model of @xcite for which a(28@xmath6m)/n(h ) @xmath47 mag@xmath48 , a(17@xmath6m)/n(h ) @xmath4 a(12@xmath6m)/n(h ) @xmath49 mag@xmath48 , and make the correction for the higher gas to dust mass ratio observed in ngc891 ( 260 , see * ? ? ?
* ) as compared with the galaxy ( 160 , see * ? ? ?
we assumed a `` mixed '' extinction scenario , where the dust is mixed with the warm h@xmath0 gas , and for which the extinction correction , c(@xmath46 ) is given by c(@xmath46 ) = @xmath50(1 - e@xmath51)@xmath40 . at the nucleus
we derive a ( beam averaged ) extinction correction for the s(0 ) line of c(28@xmath6m)@xmath41.38 , and for the s(1 ) and s(2 ) lines c(12@xmath6 m ) , c(17@xmath6m)@xmath41.71 . the extinction correction as a function of offset from the nucleus
is plotted in figure 4b .
the correction is significant for much of the inner ( r < 5 kpc ) galaxy , and larger for the s(1 ) and s(2 ) lines , so that the extinction correction will change the line ratios .
the line fluxes change only gradually over the plane of the galaxy , and the slits likely enclose nearly all of the flux perpendicular to the plane .
therefore , to find the total line luminosity , we simply integrate the line fluxes along the plane of the galaxy , interpolating in a linear manner between samplings .
for both the s(0 ) and s(1 ) lines , this amounts to integrating the observations and multiplying by the ratio of the sampling interval to the beam size along the plane : 30/22.3 = 1.35 and 30/4.7 = 6.38 for the s(0 ) and s(1 ) lines respectively .
the total line fluxes so obtained are f(s(0 ) ) @xmath52 w m@xmath53 , and f(s(1 ) ) @xmath54 w m@xmath53 . at a distance of 9.5 mpc ,
the corresponding line luminosities are @xmath55 , and @xmath56 l@xmath2 for the s(0 ) and s(1 ) lines respectively .
the s(2 ) line is not as well sampled as the s(1 ) line due to sensitivity limits , but assuming the s(2)/s(1 ) line ratio for the detected positions ( @xmath4 0.59 ) is the same throughout the galaxy , the s(2 ) line luminosity would be @xmath57 l@xmath2 .
the total observed ( and interpolated ) power for these three low - lying lines is therefore @xmath58 l@xmath2 , or @xmath59 of the total ir ( tir , 3 - 1100 @xmath6 m ) luminosity , @xmath60 l@xmath2 ( flux densities from @xcite , tir as defined in @xcite ) .
the @xmath61 low j molecular hydrogen lines are therefore @xmath4 1.5 times as bright relative to the tir continuum when compared to the sample of 57 normal galaxies surveyed by @xcite who find a typical ratio near @xmath62 with a scatter of about a factor of two .
correction for extinction before integrating along the plane ( factors of 1.16 , 1.32 , and 1.32 for the s(0 ) , s(1 ) and s(2 ) lines respectively , section 3.3 , above ) , which is bound to be more important for the edge - on ngc891 than for the typically less inclined galaxies in the @xcite sample , the total h@xmath0 luminosity is @xmath63 l@xmath2 or @xmath64 of the tir luminosity , or @xmath4 twice the typical values found in @xcite . the question of line flux at high scale heights missed by the spitzer beam is more difficult to quantify . if the scale height derived by @xcite applies the flux we observe is roughly equal to that we miss at scale heights greater than that of our beam ( @xmath4 5.6@xmath27 ) .
the missed ( high z ) emission summed over the three lines is in this case therefore @xmath58 l@xmath2 .
the high - z emission would suffer very little extinction , so the the total luminosity in all three lines would be @xmath65 l@xmath2 in the case of a 9.1@xmath27 @xcite scale height . another estimate is presented by presuming the h@xmath26 scale height is the same as that of the co(1 - 0 ) line . since the co scale height is observed along the entire galactic plane , this is the correction we prefer .
for this case , the flux exterior to our beam ( without extinction correction ) is just 16@xmath66 of the flux within our beam ( before extinction correction ) so that the total luminosity in all three lines is @xmath67 l@xmath2 .
correcting for both extinction and extra - planar emission , the h@xmath0 luminosity is therefore @xmath68 to @xmath69 of the tir luminosity .
to derive the physical properties of the emitting gas , we need to compute ratios between lines measured within different apertures . inspecting figure 3 ,
it is clear that for the most part , the flux in the three lines changes gradually at the scale of lh slit length ( 22.3@xmath27 ) along the plane of the galaxy .
the s(1 ) and s(2 ) lines are observed with the long ( 11.3@xmath27 ) axis of the sh slit perpendicular to the plane so that it matches the z - sampling of the s(0 ) beam . therefore , to take ratios among the h@xmath26 lines , we need only correct for the beam size along the galactic plane .
we do this by linearly interpolating between s(1 ) and s(2 ) samples ( taken with a 4.7@xmath27 beam along the plane of the galaxy ) , and then estimating the average intensity expected in the 22.3@xmath27 s(0 ) beam from this interpolation . to obtain the physical parameters of the gas
, we use the extinction corrected values of the fluxes contained within our spitzer beams presented in figure 5a .
the extinction corrected line ratios are presented in figure 5b .
the low lying rotational lines of h@xmath0 are optically thin and easily thermalized so that the lines and their ratios probe the temperature and column density of the emitting gas . the s(2 ) and s(0 ) lines both arise from the para species of h@xmath0 so that it in the high density limit ( n@xmath70n@xmath71(j=4 ) @xmath72 @xmath14 , @xcite ) , the i@xmath73/i@xmath74 ratio is given by the para excitation temperature t@xmath3 : @xmath75 where a@xmath76 is the einstein a coefficient for spontaneous emission of level i ( a@xmath77 = 2.75@xmath78 , a@xmath79 = 2.94@xmath80 , a@xmath81 = 4.76@xmath82 * ? ? ? * ) , @xmath83 is the frequency of transition , g@xmath84 is the statistical weight of the level ( g@xmath77=9 , g@xmath79=5 ) , and @xmath5e is the energy difference between the emitting levels (= 1172 k ) .
we plot the j=4 level excitation temperature so derived in figure 6 ( marked as s(2)/s(0 ) ) .
the excitation temperature is remarkably uniform over the disk at t@xmath3(2,0)[email protected] . between the para and ortho species ,
the line intensity ratios have the additional factor of the relative abundance , the ortho to para ratio , @xmath7 . for a high temperature ( t @xmath10 200 k ) gas in local thermodynamic equilibrium
, this ratio is given by the ratio of the statistical weights , of the two species , @xmath7 = 3 .
assuming the @xmath7 ratio is 3 , then averaged over the eleven positions for which the s(2 ) line is detected ( hereafter , the
s(2 ) region " ) , the average extinction corrected s(1)/s(0 ) and s(2)/s(1 ) line intensity ratios ( 1.92 and 0.59 ) yield level excitation temperatures of t@xmath3(1,0 ) = 125 k , t@xmath3(2,0 ) = 204 k , and t@xmath3(2,1 ) = 388 k respectively .
the level excitation temperatures are monotonically increasing with j , which is permitted if there is more than one emitting components within the beam , each with a different gas excitation temperature . for temperatures less than @xmath85k ,
the equilibrium @xmath7 ratio is less than 3 .
@xcite calculate the equilibrium @xmath7 ratio as a function of gas temperature . making use of this work , we arrive at a self consistent solution for t@xmath3(1,0 ) = 134 k with @xmath7 = 2.30 .
since the s(1)/s(0 ) line intensity ratios are fairly uniform across the galaxy , we assume this @xmath7 ratio applies everywhere , and use it to compute t@xmath3(1,0 ) . notice that t@xmath3(2,0 ) is not affected by the @xmath7 ratio , and t@xmath3(2,1 ) is much greater than 200 k , so that @xmath86 .
figure 6 shows the computed excitation temperatures across the disk of the galaxy . as we found for t@xmath3(2,0 ) , t@xmath3(1,0 ) is also remarkably uniform across the plane with a mean value of t@xmath3(1,0)=132@xmath326k .
t@xmath3(2,1 ) shows more variation , but this is largely not statistically significant . excluding the data point 6 kpc to the ne of the nucleus , the mean value is t@xmath3(2,1 ) = 372 @xmath3241k .
the position at 6 kpc ne of the nucleus is statistically higher excitation , with t@xmath3(2,1 ) = 556 + 127/-87 k. since the line emission is optically thin , the line intensity yields the gas column of the emitting level , n@xmath87 : i = h@xmath88 a@xmath89 n@xmath90 , where a@xmath89 is the einstein a coefficient for the s(n ) transition at frequency @xmath88 for n = 0,1,2 .
averaged over the eleven positions from which the s(2 ) line is detected , the column densities are n@xmath91@xmath39 , n@xmath92 @xmath39 , and n@xmath93 @xmath39 .
the total column density of molecular hydrogen is given by : n@xmath94 = ( n@xmath87/g@xmath87 ) z(t ) e@xmath95 where g@xmath87 = ( o / p)/(2j+1 ) , and z(t ) is the partition function : @xmath96 the symbols have their meanings as before .
since the level excitations are not all the same , there must be more than one warm molecular gas component .
the simplest model has two gas components , a warm " component , which will emit most of the observed s(0 ) line , and a hot " component that will emit most of the observed s(2 ) line .
both components will emit in the s(1 ) line . with a two component model
there can be no unique solution as there are 4 variables ( 2 components , each with a t@xmath3 and n(h@xmath26 ) ) , and just 3 observables ( the lines ) .
we model the line emission averaged within the
s(2 ) region " .
the minimum mass solution will have a high gas temperature for the warm component .
the largest gas temperature permitted by the data is the observed 1 - 0 excitation temperature .
for this temperature , all of the observed s(0 ) and s(1 ) line emission arises from the warm component .
therefore , there can be no self - consistent solution for all three h@xmath26 lines since the warm component is too cool to emit much in the s(2 ) line , and hot component , which will emit most of the the s(2 ) line , will also emit in the s(1 ) line . using the partition function , and the relationship between line intensity and column density ( i@xmath97 = h@xmath98@xmath97a@xmath97n@xmath99/(4@xmath100 ) )
, we have explored the gas temperature , column density parameter space for self consistent solutions , and find the minimum mass solution has gas temperatures of t@xmath101 = 127 k ( o / p ratio shifts to 2.2 ) , and t@xmath102 = 1700 k , and h@xmath26 column densities of n(h@xmath103 ) = 3.8 @xmath104 @xmath39 , and n(h@xmath105 ) = 8.2 @xmath106 @xmath39 . within this model , 75@xmath66 and 25@xmath66 of the observed s(1 )
line emission arises from the warm and hot components respectively .
the warm component emits less than 5@xmath66 of the observed s(2 ) line , while the hot component emits less than 1@xmath66 of the s(0 ) line .
the case where 25@xmath66 and 75@xmath66 of the observed s(1 ) line emission arises from the warm and hot components respectively is obtained for gas temperatures of t@xmath101 = 107 k ( o / p ratio shifts to 1.65 ) , and t@xmath102 = 465 k , and h@xmath26 column densities of n(h@xmath103 ) = 5.8 @xmath104 @xmath39 , and n(h@xmath105 ) = 3.4 @xmath107 @xmath39 . for this case ,
less than 1@xmath66 of the observed s(2 ) ( s(0 ) line ) arise from the warm ( hot ) component . for all cases ,
the warm component completely dominates ( more than @xmath108 ) the h@xmath26 column density .
we use the minimum mass self consistent model ( 75/25 split of s(1 ) line flux into warm and hot components ) to calculate the total column density in h@xmath26 line emitting molecular gas ( esssentially the warm component from above ) across the disk of ngc891 . the results , along with the column traced in co(1 - 0 ) are plotted in figure 7a . the warm molecular gas column peaks at the nucleus at a value of @xmath109 @xmath39 , and is above @xmath110 @xmath39 out to @xmath32 8 kpc from the nucleus .
figure 7b shows the ratio of the warm molecular gas to that traced in its co(1 - 0 ) line emission ( within the spitzer beam ) , i.e. the fraction of the total molecular gas that is emitting in the h@xmath26 lines .
this fraction is modest ( @xmath413% ) and nearly constant in the inner galaxy from 3 kpc ne of the nucleus to 6 kpc se of the nucleus , but rises to @xmath4 30% in the outer galaxy .
this could be a true enhancement of the warm molecular gas mass in the outer galaxy , or it could be due to a change in the conversion factor , x between co intensity and h@xmath0 column density with distance from the nucleus .
if the x factor were larger by a factor of two in the outer galaxy compared with the inner galaxy , at first glance , this could explain our apparent change in the warm / cold molecular gas fraction .
however , we do not believe this is the case , as there is a significant ne - sw asymmetry in the warm / cold molecular gas fraction apparent within the molecular ring ( r = 3 to 7 kpc ) .
averaged over the ring positions , the ne portion of the ring has twice the warm molecular gas fraction ( @xmath4 28% ) as the sw portion of the ring ( @xmath414% ) .
the total molecular gas as traced by co is very close to the same in both sides of the ring @xcite : it is the warm molecular gas column that is enhanced ( figure 7 , top ) . to cancel out this effect
, one would need to invoke an x factor twice as large for the ne ring as for the sw ring .
this enhanced warm molecular mass fraction is more likely the result of enhanced starformation in the ne ring , which is reflected in the stronger [ cii ] and [ oi ] line emission observed to the ne as well ( see figure 8) .
integrated over the galaxy , the warm molecular gas totals @xmath111 m@xmath2 , or about 16% of the total molecular gas mass as traced by co ( @xmath112 m@xmath2 ) within our spitzer beams ( which extend @xmath325.5@xmath27 of the galactic plane ) .
the power to heat large amounts of molecular gas to temperatures above a hundred degrees can be provided by far - uv ( stellar ) photons , by the dissipation of molecular gas turbulent kinetic energy in cloud - cloud collisions and on smaller scales within clouds , by x - ray photons , or by cosmic rays . for normal star - forming galaxies ,
only the first two may provide enough power to be important .
figure 8 provides insights into the source for the heating of the molecular gas . in this figure
we plot the cold " molecular gas mass ( traced in its co emission ) the warm " molecular gas mass ( traced in its h@xmath0 line emission ) , and the [ cii ] 158 @xmath6 m and the [ oi ] 63 fine - structure lines obtained with the iso lws spectrometer @xcite .
we also include weak detections of the [ oi ] 146 @xmath6 m at 6 positions along the galactic plane that we obtained through close inspection of the lws l02 scans of the line obtained within the iso data archive .
these data were calibrated as described in @xcite so as to be consistent with their [ cii ] and [ oi ] 63 @xmath6 m fluxes , and their [ nii ] 122 @xmath6 fluxes which we use below .
the cold and warm molecular gas distributions have been smoothed to the significantly coarser spatial resolution ( @xmath475@xmath27 ) of the iso lws beam for proper comparisons , and the tracers are normalized to their nuclear values ( excepting the [ oi ] lines which are normalized to their peaks ) for ease of comparision .
the cold molecular gas mass is centered on the nucleus with a significantly narrow distribution along the galactic plane than the warm molecular gas mass . in addition , the warm molecular gas mass distribution is asymetrical , with near nuclear values extending to the molecular ring , 4.6 kpc to the ne of the nucleus .
the [ cii ] line predominantly arises from photodissociation regions formed on the surfaces of molecular clouds exposed to the far - uv ( 6ev @xmath113ev ) radiation from nearby ob stars or the general interstellar radiation field .
the [ cii ] line is also an important coolant for low density ionized gas , and atomic clouds " , but most ( @xmath10 70% * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) of the [ cii ] flux from external galaxies likely arises from pdrs , so that the line is an indicator of ob star formation activity @xcite .
the very close correspondance between the [ cii ] distribution and that of the warm molecular gas is strongly suggestive of a pdr origin for the h@xmath26 line emission .
the ionization potential for o ( 13.62 ev ) is essentially the same as that of h , so that the fir [ oi ] 63 and 146 @xmath6 m line emission also arises from neutral gas within pdrs .
the [ oi ] lines have significantly higher critical densities ( @xmath4 4.7 @xmath22 10@xmath114 and 9.4 @xmath22 10@xmath115 @xmath14 for the 63 and 146 @xmath6 m lines respectively ) than that of the [ cii ] line ( @xmath4 2.8 @xmath22 10@xmath116 @xmath14 ) so that the [ oi ] lines trace denser pdrs .
the strong peaking of the [ oi]63 @xmath6 m line line on the northern molecular ring is strongly suggestive of enhanced gas densities in this region . the remarkable spatial correlation between the warm h@xmath0 mass and the [ cii ] line including the extension to the ne suggests a common origin in pdrs .
indeed , warm molecular gas is expected within pdrs .
the penetration of carbon ionizing photons into neutral gas clouds is limited by the extinction of these photons by dust to visual extinctions , a@xmath117 mag ( corresponding to an n@xmath118 @xmath39 ) , and along half this column , hydrogen is typically molecular in form @xcite .
we have argued for a pdr orgin for much of the h@xmath26 line emission . as such , one might expect a correlation between the observed h@xmath0 line emission and that of an independent tracer of pdrs , the pah 7.7 @xmath6 m emission feature ( e.g. @xcite,@xcite . in figure 9
, we compare the pah 7.7@xmath6 m emission with the h@xmath0 line emission , by plotting the h@xmath0-to - pah emission ratio along the plane of ngc891 .
the pah emission is measured from irac 8 @xmath6 m imaging data @xcite using the same apertures than those used for the h@xmath0 spectroscopy . to compute the pah emission from the 8 @xmath6 m flux , we use the same effective bandwith ( @xmath120 ) as @xcite .
the h@xmath0 line emission is the sum of the s(0 ) , s(1 ) and s(2 ) lines .
all tracers were corrected for extinction ( see section 3.3 ) .
the extinction at 7.7 @xmath6 m is similar to that at 12.3 and 17 @xmath6 m @xcite , so that the h@xmath0 to pah emission ratio is not very sensitive to the specific choice of the extinction curve . @xcite and @xcite have shown that the h@xmath0 emission of star forming galaxies is tightly correlated with the pah emission .
figure 9 extends this correlation to the distribution of both tracers within a galaxy .
the h@xmath0 to pah emission ratio is remarkably constant within the ngc891 disk , with a mean value only slightly larger than that measured for the sings galaxies .
a striking outcome of the sings data analysis is that the h@xmath0/pah emission ratio is independent of the dust heating as measured by the @xmath121 m to fir dust emission ratio .
@xcite argue that this result is easiest to understand if the h@xmath0 and dust emission come both from pdrs , and are powered by a common energy source , the uv light from young stars .
however , from their discussion it is not fully clear how this interpretation fits with the modeling of the dust spectral energy distribution in @xcite , where the pah emission comes mainly from interstellar matter in a low uv radiation field ( typically g @xmath122 ) . here
, we quantify the constraint set by the h@xmath0/pah emission ratio on the uv field . using the @xcite models available on line , we computed that the pah emission , as it appears in figure 9 , is @xmath123 where the dus - to - gas mass ratio is assumed to be galactic and q@xmath124 is the fraction of the dust mass in pahs @xcite .
the default value of q@xmath124 corresponds to the median value derived by @xcite for metallic galaxies in the sings sample .
the model output may be combined with the observed value of h@xmath0/pah ( @xmath4 0.012 ) , to estimate the effective radiation field , g , within the warm h@xmath0 gas . for ngc 891
, we find l(h@xmath0)/m(h@xmath0,warm ) @xmath4 1/20 , so that putting this all together we have : @xmath125 here we have left in an explicit scaling factor for the warm molecular gas mass . clearly , this calculation only provides a rough upper limit for g , since it assumes that the pah emission arises exclusively from pdrs containing the warm h@xmath0 gas , while some pah emission is expected from the h i and cold h@xmath0 gas .
however , within the current understanding of pah emission , the h@xmath0 to pah comparison therefore suggests that the bulk of the warm h@xmath0 mass ( the s(0 ) emitting gas ) is located in pdrs exposed to modest far - uv radiation fields .
we use this as a constraint in our pdr models that we develop below .
it is important to note that this constraint on g only applies for the gas that radiates in the s(0 ) and s(1 ) lines , not the gas radiating in the s(2 ) line .
the s(2 ) line emission could arise from a much higher g environment without increasing the pah flux very much since the mass of this component will be so small .
@xcite have made available a grid of pdr model outputs , which include the expected rotational line emission from h@xmath0 as a function of g , the strength of the local far - uv ( 6 @xmath126 ev ) radiation field normalized to the habing field ( g@xmath127 erg s@xmath40 @xmath39 ) , and the cloud number density n. the model computes the local heating and cooling , chemistry and radiative transfer as a function of depth from the cloud surface into its dust - shielded core .
diagnostic line ratios and line intensities appropriate for a single face - on slab geometry are posted on - line , and available interactively in the photodissociation region toolbox ( pdrt ) " .
table 2 lists our observed h@xmath0 s(0 ) , s(1 ) and s(2 ) luminosities and those of the [ cii ] , and [ oi ] lines integrated over the galaxy .
we estimate that the errors for the h@xmath0 luminosities are roughly given by their corrections for extinction , and high latitude flux missed by the spitzer beam .
figure 10 ( left ) plots our observed galactic average h@xmath0 line ratios as a function of g and n from pdrt .
the model results do not give a good fit to the ngc891 line ratios .
the best fit is for high density ( n@xmath128 @xmath14 ) and moderate far - uv radiation fields ( g @xmath4 200 to 400 ) .
figure 10 ( center ) plots the [ oi ] , [ cii ] and ( [ oi ] 63 @xmath6 m + [ cii])/fir continuum ratios as averaged over the galaxy onto a pdr diagnostic diagram from pdrt . for this fir data set
we make corrections to the observed luminosities of our tracers .
the first correction is for optical depth and geometry .
the pdrt models are appropriate for face - on single slab geometry .
since the iso beams encompass regions @xmath4 2.5 kpc in size at the distance to ngc891 , there is an ensemble of clouds within our beam . for a spherical cloud ,
the projected surface area is only 1/4 the total surface area , so that a beam filled with spherical clouds will have 4 times the intensity as a unit filling factor single slab if the lines are optically thin .
if all the tracers are optically thin , taking the line ratios eliminates this geometric factor .
although it is almost certainly true that all the h@xmath0 lines are optically thin , optical depth may be an issue for the fir lines .
in fact , the 63 @xmath6 m [ oi ] is most likely optically thick @xcite in star formation regions , so that the observed [ oi ] line intensity should be corrected for geometry , multiplying by as much as a factor of four for high optical depth in the line in spherical geometry . for our study
, we make a more modest correction , multiplying the 63 @xmath6 m line by a factor of two .
we take the [ oi ] error boundary to be 50@xmath66 due to this factor of two correction , and note that this error bound will encompass geometry effects up to a factor of 3 .
the [ cii ] and [ oi ] 146 @xmath6 m lines are likely optically thin , or thinish as is the fir continuum , so we make no optical depth correction to their fluxes .
next , we correct the observed [ cii ] line emission for that fraction expected to arise from diffuse ionized gas .
this correction is perhaps best done by comparing the observed [ cii ] line flux to that of the ground state fine structure lines of n@xmath8 at 205 and 122 @xmath6 m . with an ionization potential of 14.5 ev ,
these lines only arise from within ionized gas regions , and one can show that the [ cii]/[nii ] line ratios are a measure of the observed fraction of the [ cii ] line that arises from the ionized interstellar medium ( cf . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we use the [ nii ] 122 @xmath6 m lines observed by the iso lws @xcite , averaged over the plane of the galaxy .
the [ nii ] 122@xmath6 m line to [ cii ] line ratio in ngc891 is @xmath41:9 : somewhat smaller than that of the milky way galaxy ( 1:6 , * ? ? ?
assuming galactic c / n abundance ratios the line ratio indicates that between 50% ( low density limit ) and 4% ( high density limit ) of the observed [ cii ] comes from ionized gas regions @xcite . here
we will assume 27% of the [ cii ] line arises from hii regions as found for the low density ( n@xmath129@xmath4 30 @xmath14 ) gas in the carina nebula@xcite .
we take the error bound for the [ cii ] luminosity to be equal to the fraction ( 27@xmath66 ) we have subtracted off . finally , not all of the fir continuum luminosity from galaxies will arise from dense pdrs . a significant fraction can arise from the diffuse ism . @xcite and @xcite have fit warm " ( pdr ) and cool " ( diffuse ) fractions of the total fir luminosity from ngc891 .
the warm dust likely arises from h ii regions and pdrs , while the cool dust arises from the cold interiors of molecular clouds and the diffuse ism @xcite .
we therefore take @xmath450@xmath66 of the total fir luminosity of ngc891 to arise from pdrs . finally , we correct l@xmath130 to l@xmath131 which is appropriate for the pdrt models ( see footnote to table 2 ) . considering these corrections ,
we take the error bound for the fir luminosity to be 50@xmath66 . putting the fir line and continum corrections together ,
the pdr cooling line / cooling continuum ratio integrated over ngc891 is ( [ oi ] + [ cii])/fir @xmath41.7% ( table 2 ) .
let us now examine the overall properties of ngc891 , integrated along the galactic plane .
taken within themselves , the s(2)/s(0 ) and s(1)/s(0 ) line ratios ( 1.08 and 1.69 respectively ) indicate g @xmath4200 - 400 , and n @xmath366 @xmath22 10@xmath132 @xmath14 . however , the [ oi ] 146@xmath6m/[cii ] , [ oi ] 63@xmath6m/[cii ] and ( [ oi]+[cii])/fir ratios indicate lower far - uv fields with best fit near g@xmath4100 , and a significantly lower density with best fit near n@xmath4 3@xmath133 ( figure 9 , center ) . the solution spaces for
the h@xmath26 and fir line models do not overlap .
the closest approach for the two models is obtained for similar gs ( @xmath4 100 to 200 ) , but the fir lines require a lower gas density ( n@xmath4 10@xmath115 to 4 @xmath134 @xmath14 , as g ranges from 100 to 200 ) than the h@xmath0 lines ( n@xmath135 @xmath14 ) . if we imagine the pdrs to be the surfaces of externally illuminated molecular clouds , then since the h@xmath0 lines arise from deeper within the pdr than the fir lines , it is likely they will arise from denser gas regions , so perhaps this is an acceptable solution .
if so , then the beam filling factors of the two components should match as well .
for the fir component , the average [ cii ] line intensity ( less 27% for ionized gas ) referred to an 11@xmath27 region along the plane of the galaxy is @xmath136 erg s@xmath40 @xmath39 sr@xmath40 . at the closest approach solution between the h@xmath26 and fir models ( n@xmath137 @xmath14 and g@xmath4200 ) , the [ cii ] intensity predicted by pdrt
is @xmath138 erg s@xmath40 @xmath39 sr@xmath40 so that the beam filling factor for the fir lines is @xmath139 @xmath41.3 . for n@xmath140 ,
g@xmath4200 , pdrt predicts a s(0 ) line intensity @xmath141 erg s@xmath40 @xmath39 sr@xmath40 .
the extinction corrected value within our 11@xmath27 beam averaged over the plane of the galaxy is @xmath142 erg s@xmath40 @xmath39 sr@xmath40 , so that the beam filling factor is @xmath143 @xmath4 1.7 , in fair agreement with the fir value .
the h@xmath26 lines , the fir lines , and the fir continuum are consistent with pdr models but only if there is a rather large ( factor of @xmath410 ) density gradient between the ( denser ) h@xmath0 emitting region and the ( less dense ) fir line emitting regions .
in addition , the requisite far - uv field strength is significantly larger than the value we derive using the observed pah distribution in section 4.3.2 above .
these difficulties motivate us to investigate the possibility that a fraction of the observed h@xmath0 line emission arises from the dissipation of turbulence in molecular clouds .
this possibility has been proposed to account for diffuse h@xmath0 line emission in the milky way @xcite and the galaxy - wide shock in stephan s quintet @xcite .
the s(1)/s(0 ) line ratio is fully consistent with the far - uv fields and densities as derived from the fir lines and continuum ( see figure 10 , left ) .
what drives the h@xmath0 solution is the s(2 ) line , which is a factor of six too bright .
the c - shock models of @xcite and , more recently @xcite are in good agreement , and indicate that the s(2 ) line is brighter than the s(1 ) line for v@xmath144 @xmath36 15 km s@xmath40 .
for example , using the @xcite model , for shock velocities , v@xmath14520 to 30 km s@xmath40 , with preshock density n@xmath146 = n(h)+2n(h@xmath0)= 2@xmath2210@xmath132 @xmath14 , the s(2 ) line is @xmath4 1.6 @xmath22 as bright as the s(1 ) line , and 75 times brighter than the s(0 ) line .
if 80% of the observed s(2 ) line flux were to arise from such c - shocks , and the remainder from pdrs , then 30% of the s(1 ) line would also arise from shocks , and the pdr line ratios would shift to s(2)/s(1 ) @xmath40.17 , and s(2)/s(0)@xmath4 0.22 .
the allowed values for g and n for this pdr plus shocks solution are shaded blue in figure 10 ( right ) .
clearly there is now a large region of overlap between the pdr solution for the fir lines , and that for the h@xmath26 lines .
this solution space is centered at about g@xmath460 @xmath32 40 , n@xmath4 5 + 5/-3 @xmath134 @xmath14 ( figure 10 , center ) .
the revised filling factors are reasonably self - consistent with @xmath139 @xmath4 1.9 , and @xmath143 @xmath4 2.8 . with v@xmath147 25
km s@xmath40 , there is significant cooling in other molecular hydrogen lines : the s(3 ) and s(4 ) lines are predicted to be 4@xmath47 and 4 times brighter respectively than the s(2 ) line .
the total luminosity in the molecular hydrogen lines from the shocked gas would be very large : l@xmath148 l@xmath2 .
for a 25 km s@xmath40 c - shock , the common cooling lines of [ oi ] and co are more than 100 and 15 times less important energetically than those of h@xmath0 , and h@xmath0o is predicted to have a total cooling @xmath4 that of h@xmath0 @xcite .
a similar analyis of the region centered on the north - east ring has a solution centered on modestly enhanced excitation , g@xmath14980 , n@xmath150 @xmath14 , and requires somewhat higher area filling factors @xmath139 @xmath4 2.7 , and @xmath143 @xmath4 3.6 to account for the enhanced [ oi ] 63 @xmath6 m line emission there .
leaving aside the question of the h@xmath0 excitation , we can assess the contribution of mechanical energy dissipation to the h@xmath0 emission , from the point of view of energetics .
is the dissipation rate of the interstellar medium turbulent kinetic large enough to provide a significant contribution to the h@xmath0 emission ?
various groups have discussed the dissipation of turbulence using numerical magneto - hydrodynamic simulations ( * ? ? ?
* ; * ? ? ?
* e.g. ) . as in @xcite
, one can rearrange equation ( 7 ) from @xcite to parameterize the energy released by the dissipation of turbulence per unit mass in terms of the dispersion of the turbulent velocity , v@xmath151 , and the scale length of energy injection , @xmath152 : @xmath153 the default values for v@xmath151 and @xmath152 correspond to the upper end of the line width - size relation for giant molecular clouds in the galaxy @xcite .
we invoke shocks to account for 80% of the s(2 ) luminosity , and 25% of the s(1 ) line luminosity . by dividing the sum of these line
luminosiites @xmath154 , by the molecular mass @xmath155 , derived from the co(1 - 0 ) luminosity , we obtain @xmath156 .
therefore , the expected turbulent dissipation rate is greater than the luminosites of the lines , so that a significant contribution from turbulent energy dissipation is likely . in summary ,
the observed h@xmath26 line emission from ngc 891 is well fit by a model that combines a low excitation pdr ( g @xmath4 60 , n@xmath45@xmath157 10@xmath116 @xmath14 ) with dissipation of turbulent energy within molecular clouds .
the ( adjusted ) [ oi ] ( 63 and 146 @xmath6 m ) and [ cii ] 158 @xmath6 m fine structure line fluxes , ( half ) of the observed fir continuum flux , the h@xmath26 s(0 ) line flux and 70@xmath66 of the s(1 ) line flux arise from the pdr component , while the h@xmath26 s(2 ) line and 30@xmath66 of the s(1 ) line flux arise from gas heated by turbulent dissipation .
it is interesting to note that based on observations of the h@xmath0 lines along a line of sight within the milky way galaxy , @xcite reached much the same conclusion : the s(0 ) line emission may be accounted for by uv heated h@xmath0 in the diffuse interstellar medium and low excitation pdrs , but the uv emission can not be the unique energy source , because their pdr models fall short of the observed s(1 ) and s(2 ) lines by large factors ( @xmath42 - 10 ) .
as we outline here , they propose that the s(1 ) and s(2 ) h@xmath0 line emission is powered by the localized dissipation of turbulent kinetic energy , albeit in a low density gas ( n@xmath158 ) within the diffuse interstellar medium .
the presence of a diffuse component of h@xmath0 line emission has been confirmed by additional galactic observations @xcite .
it should be clear that the solutions presented here are by no means unique .
on galactic scales there will certainly be a range of pdrs along the line of sight with varying degrees of excitation , that could mimic the observed line ratios , perhaps negating the need to invoke shocks .
furthermore , a single parameter shock model is also simplistic : there will very likely be a superposition of shocks of various parameters only any given line of sight .
what we present here is intended to be only a representative solution .
we have mapped the distribution of the s(0 ) , s(1 ) , and s(2 ) pure rotational lines of h@xmath26 along the plane of the edge - on galaxy ngc891 .
the lines are remarkably bright . after making an extinction correction ,
the total line flux from the three transitions is @xmath159 l@xmath2 or @xmath40.09% of the tir luminosity of the galaxy . unlike the iso results that suggest a change in the s(1)/s(0 ) line intensity ratio in the outer regions of the galaxy
, we find the line ratios are nearly constant along the plane of the galaxy indicating uniform excitation of the gas . in particular , there is no reason to invoke a massive cold component of the molecular gas in the outer galaxy as the iso results appear to require . through analytical modeling
, we find the bulk of the mass traced in its h@xmath0 line emission has an excitation temperature near 125 k. this `` warm '' molecular component is a substantial fraction of the total molecular gas mass : about 10 to 15% in the inner ( 3 kpc ne @xmath35 r @xmath35 6 kpc sw ) galaxy rising to more than 30% in the outer regions .
integrated over the galaxy , m@xmath16 @xmath4 16% m@xmath18 .
we compare our h@xmath0 observations with detailed pdr models to show that the observed line intensities and ratios are consistent with high density ( n@xmath160 @xmath14 ) molecular clouds exposed to moderate far - uv fields ( g@xmath4200 - 400 ) .
we refine our pdr modeling by using available iso [ oi ] and [ cii ] data , which indicate smaller far - uv fields ( g@xmath45 - 300 ) , and lower gas densities ( n@xmath161 @xmath14 ) .
the solutions spaces for the h@xmath0 and fir line pdrs do not overlap , but a reasonably close solution is obtained for g@xmath4 200 , and a large ( factor of ten ) density gradient between the fir emitting gas and the h@xmath0 emitting gas
. this solution may be possible if the clouds are externally illuminated , but the high field invoked by this solution is in contradiction to the more modest far - uv fields ( g@xmath162 ) one would expect by comparing the observed h@xmath0 line luminosity to that of the 7.7 @xmath6 m pah features .
a consistent solution is obtained within a two component model .
the first component is a low excitation pdr , with g @xmath4 60 , and the second component is heated by the dissipation of turbulent kinetic energy .
we find a very good fit for the h@xmath0 and fir lines and continuum if all the observed s(0 ) line emission arises from the pdr component , and 80@xmath66 of the s(2 ) line arises from the turbulent component .
the s(1 ) line emission is split 70% into the pdr component , and 30% into the turbulent component .
this solution is only representative of a range of solutions that could be obtained by allowing more pdr and/or shock components with different excitation parameters .
this is the first analysis of the h@xmath0 rotational line emission from a galaxy together with the fir lines and continuum in the context of pdr models .
a complication with this modeling is the relatively large beams and modest sensitivies of the iso spectrometers for detection of the fir lines .
these issues will become much less important in the near future with the arrival of fir data from the 3 m class stratospheric observatory for infrared astronomy ( sofia ) and the herschel space telescope spectrometers and photometers .
we thank an anonymous referee for many inciteful comments that led to a much improved version of this manuscript .
this work is based in part on observations made with the spitzer space telescope , which is operated by the jet propulsion laboratory , california institute of technology , under nasa contract 1407 .
support for this work was provided by nasa through contract numbers 1257184 and 1311091 issued by jpl / caltech .
vc acknowledges partial support from the eu tok grant 39965 and fp7-regpot 206469 .
rcccccc 26 & 2:22:44.10 & + 42:26:2.450 & 5.65 & @xmath351.40 & @xmath351.24 & @xmath351.44 + 25 & 2:22:43.35 & + 42:25:33.85 & 5.53 & @xmath351.47 & @xmath351.39 & @xmath352.10 + 24 & 2:22:42.37 & + 42:25:06.67 & 4.97 & @xmath351.74 & @xmath351.37 & [email protected] + 23 & 2:22:41.44 & + 42:24:40.09 & 4.42 & @xmath351.67 & @xmath351.58 & [email protected] + 22 & 2:22:40.78 & + 42:24:07.65 & 3.87 & [email protected] & [email protected] & [email protected] + 21 & 2:22:39.83 & + 42:23:39.90 & 3.32 & @xmath351.89 & [email protected] & [email protected] + 20 & 2:22:38.38 & + 42:23:13.93 & 2.76 & [email protected] & [email protected] & [email protected] + 19 & 2:22:37.59 & + 42:22:46.22 & 2.21 & [email protected] & [email protected] & [email protected] + 18 & 2:22:36.47 & + 42:22:18.77 & 1.66 & [email protected] & [email protected] & [email protected] + 17 & 2:22:35.47 & + 42:21:50.29 & 1.11 & [email protected] & [email protected] & [email protected] + 16 & 2:22:34.67 & + 42:21:21.43 & 0.55 & [email protected] & [email protected] & [email protected] + 1 & 2:22:33.41 & + 42:20:56.88 & 0.00 & [email protected] & [email protected] & [email protected] + 2 & 2:22:32.37 & + 42:20:25.85 & -0.55 & @xmath351.58 & [email protected] & [email protected] + 3 & 2:22:31.02 & + 42:19:59.96 & -1.11 & [email protected] & [email protected] & [email protected] + 4 & 2:22:30.16 & + 42:19:31.63 & -1.66 & [email protected] & [email protected] & [email protected] + 5 & 2:22:29.15 & + 42:19:03.34 & -2.21 & [email protected] & [email protected] & [email protected] + 6 & 2:22:28.07 & + 42:18:35.81 & -2.76 & [email protected] & [email protected] & [email protected] + 7 & 2:22:26.99 & + 42:18:08.74 & -3.32 & @xmath352.70 & [email protected] & [email protected] + 8 & 2:22:25.98 & + 42:17:41.93 & -3.87 & @xmath352.19 & [email protected] & [email protected] + 9 & 2:22:24.88 & + 42:17:14.69 & -4.42 & @xmath352.40 & [email protected] & [email protected] + 10 & 2:22:23.74 & + 42:16:46.54 & -4.97 & @xmath351.41 & [email protected] & [email protected] + 11 & 2:22:22.70 & + 42:16:19.70 & -5.53 & @xmath351.64 & @xmath351.05 & @xmath352.24 + 12 & 2:22:21.65 & + 42:15:52.55 & -6.08 & @xmath351.44 & @xmath351.07 & @xmath351.66 + lccc h@xmath0 s(0 ) & 5.44@xmath163 & 7.18@xmath163 & 24% + h@xmath0 s(1 ) & 8.94@xmath163 & 1.32@xmath164 & 24% + h@xmath0 s(2 ) & 5.26@xmath163 & 7.80@xmath164 & 32% + 63 @xmath6 m & 4.67@xmath164 & 9.34@xmath164 & 50% + 146 @xmath6 m & 4.74@xmath163 & 4.74@xmath163 & 50% + 158 @xmath6 m & 1.40@xmath165 & 1.02@xmath164 & 37% + l@xmath130 & 1.51@xmath166 & 1.13@xmath166 & 30% + | we have studied the molecular hydrogen energetics of the edge - on spiral galaxy ngc891 , using a 34-position map in the lowest three pure rotational h@xmath0 lines observed with the spitzer infrared spectrograph .
the s(0 ) , s(1 ) , and s(2 ) lines are bright with an extinction corrected total luminosity of @xmath1 l@xmath2 , or 0.09% of the total - infrared luminosity of ngc891 .
the h@xmath0 line ratios are nearly constant along the plane of the galaxy we do not observe the previously reported strong drop - off in the s(1)/s(0 ) line intensity ratio in the outer regions of the galaxy , so we find no evidence for the very massive cold co - free molecular clouds invoked to explain the past observations .
the h@xmath0 level excitation temperatures increase monotonically indicating more than one component to the emitting gas .
more than 99% of the mass is in the lowest excitation ( t@xmath3 @xmath4125 k ) `` warm '' component . in the inner galaxy ,
the warm h@xmath0 emitting gas is @xmath415% of the co(1 - 0)-traced cool molecular gas , while in the outer regions the fraction is twice as high .
this large mass of warm gas is heated by a combination of the far - uv photons from stars in photo - dissociation regions ( pdrs ) and the dissipation of turbulent kinetic energy . including the observed far - infrared [ oi ] and [ cii ] fine - structure line emission and far - infrared continuum emission in a self - consistent manner to constrain the pdr models , we find essentially all of the s(0 ) and most ( 70% ) of the s(1 ) line arises from low excitation pdrs , while most ( 80% ) of the s(2 ) and the remainder of the s(1 ) line emission arises from low velocity microturbulent dissipation . |
in a recent paper @xcite it has been shown that a 2d spin - orbit coupled electron - doped semiconductor , in proximity to a bulk @xmath0-wave superconductor and an externally induced zeeman splitting , can support a topological superconducting phase with majorana fermion modes at vortex cores and sample edges i.e. at order parameter defect locations .
it has also been realized from dimensional reduction that the edge majorana modes in a sample with rectangular geometry ( of width @xmath1 ) turn into localized _
majorana modes in the corresponding 1d geometry ( @xmath2 ) and , further , in 1d the excitation gap above the majorana modes ( mini - gap ) should scale as the induced superconducting gap @xmath3 ( thus avoiding the problem of a tiny mini - gap , @xmath4 mk , as in 2d @xmath5-wave superconductors ) .
the end majorana modes in the 1d geometry should be experimentally visible in local zero - bias tunneling experiments . these results ,
numerically confirmed soon thereafter , @xcite have appeared with analytical and numerical details for both 1d and 2d semiconductors in ref .
it has also been pointed out @xcite that the required zeeman splitting in 1d can be induced by a magnetic field parallel to the adjacent superconductor ( thus avoiding the problem of orbital effects @xcite ) , and the non - abelian character of the end majorana modes can be probed by josephson experiments .
very recently it has been shown that the generic luttinger hamiltonian applicable to the _ hole - doped _ nanowires also supports end majorana modes @xcite in a manner similar to its electron - doped counterpart .
the 1d electron- or hole - doped wires in the ts state can be arranged in a quasi-1d network geometry @xcite to test non - abelian statistics @xcite and perform topological quantum computation ( tqc ) @xcite in the bravyi - kitaev ( bk ) scheme . @xcite in principle , tqc in the bk scheme is also possible with the 2d semiconductor ts states using majorana fermion interferometry @xcite analogous to that in the @xmath6 fractional quantum hall ( fqh ) states , @xcite chiral @xmath7-wave superconductors , @xcite and ts states on the surface of topological insulators .
@xcite efforts to realize a semiconductor ts state with majorana fermions in proximity to @xmath0-wave superconductors are currently underway in many laboratories world - wide , concentrating on both 1d semiconducting nanowires and 2d semiconductor heterostructures in close proximity to a regular bulk @xmath0-wave superconductor ( e.g. al , nb ) . in view of the ongoing experimental efforts on semiconductor ts states ,
it is important to understand the effects of disorder from the semiconductor surface on the realizability of the ts state .
( in recent works @xcite it has been established that the disorder residing in the adjacent bulk superconductor has negligible effect on the semiconductor ts state , and therefore in this paper we will ignore the effects of disorder from the bulk superconductor . ) this is an especially important question @xcite because the ts states in semiconductors explicitly break the time reversal ( tr ) symmetry and it is known that the superconducting quasiparticle gap in such systems , unlike that in tr - invariant superconductors , @xcite is suppressed by disorder . below we will measure the extent of tr breaking in the semiconductor ts state ( in both 1d and 2d ) by the ratio @xmath8 , where @xmath9 is the zeeman energy in the ts state and @xmath10 gives the typical spin - orbit energy scale ( with @xmath11 as the rashba spin - orbit coupling constant ) . we will show that the disorder suppression of the superconducting quasiparticle gap from its clean value increases with increasing values of @xmath12 .
nonetheless , by giving explicit numbers we show that a reasonable ts state quasiparticle gap can be experimentally achievable in both electron- and hole - doped systems even in the presence of realistic disorder . for electron - doped wires
we find that with a mobility of @xmath13 @xmath14/v - s a measurable robust ts state gap of @xmath15 mk is achievable .
we also find that for hole - doped wires with a significantly larger spin - orbit energy @xmath16 mev , @xcite a larger gap of @xmath17 k is achievable for a mobility of @xmath18 @xmath14/v - s . in what follows , by @xmath19
we denote the quasiparticle gap ( pair potential ) in the ( proximity - inducing ) bulk superconductor . for the semiconductor ( 2d or 1d )
, @xmath11 is the rashba spin - orbit coupling constant , @xmath20 is the effective mass ( of electrons or holes ) , @xmath9 is the zeeman splitting required for the ts state , @xmath21 is the proximity - induced pair potential , @xmath3 is the quasiparticle gap at the fermi surface , @xmath22 is the fermi momentum ( @xmath23 ) , and @xmath24 is the effective intrinsic semiconductor disorder scattering rate . in the semiconductor ,
the combination of the zeeman splitting and the spin - orbit coupling results in a fermi - surface quasiparticle gap @xmath3 , which differs from the proximity - induced pair potential @xmath21 , and thus we have three distinct superconducting gaps @xmath19 , @xmath3 and @xmath21 to consider in this problem . in sec .
ii we first give our starting hamiltonian for the semiconductor which includes a proximity - induced pair potential , an externally applied zeeman splitting and a disorder potential .
we then calculate the dependence of the disordered ts state quasiparticle gap on the semiconductor disorder scattering rate @xmath24 ( eq .
[ scbornsol ] ) . in sec .
iii , we review the two systems that we will apply our results to , namely , electron and hole - doped semiconductors . in section iv
, we use the results of sec .
ii and sec .
iii to give explicit realistic values for the various parameters for a stable ts state in electron - doped semiconductors .
this is followed by the case of hole - doped semiconductors , where again using explicit values of the parameters we show that a stable ts state is possible .
we find that the ts state gap in the hole - doped case can be much larger and more robust to disorder effects than in its electron - doped counterpart due to larger values of @xmath11 and @xmath20 .
we conclude in sec .
v with a summary and an outlook with a detailed qualitative discussion on why the key problem of choosing the optimal materials combination for realization of the ts state is experimentally important .
the bdg hamiltonian in real - space for spin-@xmath25 electrons with spin - singlet pairing potential is written as @xmath26 where @xmath21 is a real , constant , @xmath0-wave spin - singlet pairing potential in the semiconductor which is proximity induced from an adjacent superconductor , @xmath27 is the non - interacting part of the semiconductor hamiltonian with disorder and @xmath28 is the chemical potential . here
@xmath29 are the pauli matrices . for a disordered semiconductor , with spatially localised ( i.e. short - ranged ) spin - independent disorder,@xmath30
can be written as @xmath31 where @xmath32 is the local potential induced by disorder . in the limit of weak disorder and weak pairing where both @xmath32 and @xmath21 are smaller than the inter - band spacing of the spin - orbit bands in the semiconductor
, it is convenient to work in the basis of bloch eigenstates @xmath33 of @xmath34 . in particular , in the case where a single band , say @xmath35 , crosses the fermi - surface ( which is necessary for realizing a topological superconducting state @xcite ) , we can calculate the low - energy spectrum of the hamiltonian from an effective hamiltonian which is obtained by taking matrix elements of the bdg hamiltonian in eq .
[ bdgreal ] with respect to the nambu spinors @xmath36 where @xmath37 with @xmath38 being the complex conjugation operator . in this approximation ,
the translationally invariant part of the bdg hamiltonian has the form @xmath39 where we have suppressed the band - index @xmath35 , since we are restricted to a single band and the term @xmath34 has been replaced by its eigenvalue @xmath40 in the relevant band . in the electron - doped case , @xmath40 includes effects on the dispersion of the parameters @xmath9 and @xmath11 ( see eq .
[ rashba ] ) while in the hole - doped case , it additionally contains information about the luttinger parameters ( see eq .
[ luttinger1 ] ) .
we note that , in spite of its apparent complexity involving the superconducting proximity effect , spin - orbit coupling , zeeman splitting , and disorder , we are still dealing here with an effective exactly solvable one - particle quantum problem . for weak superconducting pairing and disorder scattering ,
the ts quasiparticle gap is determined by the energies and wave - functions near the fermi energy @xmath28 .
in particular , we will approximate the dispersion around the fermi energy by @xmath41 where @xmath42 is the fermi - velocity . moreover , since the relevant states are near the fermi wave - vector ( i.e @xmath43 ) , we will assume that matrix elements such as @xmath44 depend only on the directions of the momenta @xmath45 and @xmath46 . to calculate the matrix elements of @xmath47
, we will assume that the bloch eigenstates have a simple decomposition as @xmath48 this assumption is valid as long as all perturbations involve momentum transfers that are smaller than a bloch vector . for electrons
@xmath49 is simply the electron - spin .
when applied to holes @xmath49 would represent the 4-component pseudo - spin degree of freedom associated with the luttinger model .
@xcite in this paper we will consider both 1d and 2d systems , and these cases will turn out to be closely related . in the two - dimensional case , we will restrict ourselves to rotationally symmetric systems with circularly symmetric fermi surfaces where @xmath50 . using the azimuthal symmetry of the bloch hamiltonian corresponding to @xmath51
, we can define a hermitian matrix @xmath52 , which generates @xmath53-axis rotations so that @xmath54 where @xmath55 is chosen so that @xmath56 is a single valued function of @xmath57 .
since kramer s theorem requires @xmath52 to have half - integer eigenvalues @xmath58 .
for one - dimensional systems , such as electron- or hole - doped semiconducting nanowires , we will consider systems with inversion symmetric fermi points @xmath59 . in this case , the relevant states at the fermi - level are @xmath60 .
the 2d systems with rashba spin - orbit coupling will turn out to have results that are solely dependent on the wave - functions only at @xmath61 , which are identical in form to the states @xmath60 relevant for the 1d case .
therefore , these two cases are closely related and for results such as the quasiparticle gap in eqs .
[ eq : x ] and [ scbornsol ] , we will only state the result for the 1d case and imply a similar result in 2d , which can be derived easily .
the superconducting pairing potential @xmath62 in the projected hamiltonian in eq . [ h0 ] is the matrix element of @xmath21 in eq .
[ bdgreal ] between states near the fermi - level , which is written as @xmath63 using eq .
[ ket2d ] in the rotationally symmetric 2d case , @xmath62 simplifies to @xmath64 where @xmath57 is the angle of the wave - vector @xmath45 in the @xmath65 plane .
here we have assumed that @xmath52 , which transforms @xmath66 , is odd under time - reversal symmetry .
the ts quasiparticle gap for the hamiltonian eq .
[ h0 ] , in the absence of disorder ( i.e. @xmath67 ) , is given by @xmath68 the disorder scattering , induced by the potential @xmath32 in eq .
[ hsm ] , can be projected into states near the fermi energy in a similar way as @xmath69 where @xmath70 . for spatially uncorrelated white noise disorder ( i.e. @xmath71 ) , and using the simple form for the bloch functions eq .
[ bloch ] , the disorder propagator @xmath72 can be written as @xmath73 the disorder strength @xmath74 is related to the mean - scattering time ( i.e. momentum relaxation time ) @xmath75 which in turn can be determined experimentally from mobility measurements using the equation @xmath76 where @xmath20 is the transport effective mass and @xmath77 is the density of states at the fermi surface . note that we have dropped a matrix - element factor in determining the scattering time in eq .
[ tausm ] .
this is partly to account for multi - band effects that we are not directly accounting for in this paper .
since we are using mobility only as a qualitative estimator of the scattering , the combination of eqs .
[ tausm ] , [ musm ] provide a good estimate of the disorder scattering @xmath74 .
although eqs .
12 and 13 strictly apply only in the case of white noise disorder arising from short - ranged impurity scattering , we assume that they remain valid for long - range disorder as well in the presence of random charged impurity scattering since carrier screening of the impurity potential should render the coulomb disorder into an effective short - range disorder .
in addition , our use of @xmath78 to operationally characterize the semiconductor disorder should be qualitatively valid for any kind of impurity potential . while we have made a single fermi surface approximation in eq .
[ h0 ] , this is more for the sake of simplicity . in the next paragraph , we will comment on how our results generalize to the multiple fermi surface case , as for example in the case of multi - band occupancy in the semiconductor nanowire where the fermi level could lie in some high sub - band rather than the ground 1d sub - band .
@xcite the topological superconducting quasiparticle gap in the presence of a disorder potential may be determined by calculating the disorder averaged green function within the born approximation.@xcite the disorder averaged green function can be calculated from the self - energy @xmath79 using the dyson equation @xmath80 where the green function for the clean system is given by @xmath81 the self - energy @xmath82 is approximated within the self - consistent born approximation@xcite as @xmath83 where @xmath84 is the disorder propagator from eqs .
[ vmat ] , [ disordervertex ] , @xmath85 are nambu indices , and for brevity of notation we have dropped the @xmath86 dependence of @xmath79 and @xmath87 in this equation .
the self - energy within the first - order born approximation is obtained by replacing @xmath87 by @xmath88 in eq .
[ sigma ] .
the disorder calculation for the case of multiple fermi - surfaces follows in an analogous way . since momentum is a good quantum number in eq .
[ dyson ] , eq .
[ dyson ] does not couple the multiple fermi - surfaces and therefore can be solved for each fermi surface separately . the self - energy term eq .
[ sigma ] in principle has contributions from all fermi surfaces .
however , it is dominated by the contribution from the fermi - surface with the smallest gap .
therefore , our calculation , which focuses on the fermi surface with the smallest gap , is expected to yield qualitatively correct results .
solving the dyson equations in the first - order born approximation in the 1d and 2d cases , as discussed in appendix [ firstorderborn ] , we obtain the disorder averaged ts quasiparticle gap @xmath89.\label{eq : x}\ ] ] note that the disorder renormalization of the gap vanishes in the time - reversal symmetric limit since in that case @xmath47 and @xmath90 are kramer s pairs and @xmath91 . the result eq .
[ eq : x ] gives the reduction in the quasiparticle gap due to weak disorder within the first - order born approximation .
the dyson equations , eqs .
[ dyson ] , [ sigma ] , can in fact be solved for arbitrary disorder strengths within the self - consistent born approximation as explained in appendix .
[ selfconsistentborn ] .
the resulting ts state quasiparticle gap in the presence of disorder is found to be @xmath92 where @xmath93 .
note that the gap @xmath94 within the first - order born approximation in eq .
[ eq : x ] is not simply obtained by taking the limit of small @xmath95 in the self - consistent born result eq .
[ scbornsol ] .
this is because the born approximation is a series expansion in terms of the green function , which has a pole at @xmath96 , so that the expansion parameter in the series expansion diverges ( at any value of @xmath95 ) at the relevant point and one can not expect the lowest order approximation to match the self - consistent result . despite , this technical point ,
the first - order and self - consistent born approximation results , eq . [ eq : x ] and eq . [ scbornsol ] are qualitatively similar at small values of the parameter @xmath95 .
the above equation is valid as long as the expression under the square root is positive .
in fact , the quasiparticle gap @xmath94 closes for sufficiently strong disorder ( or sufficiently small @xmath97 ) so that , by solving for @xmath98 in eq . [ scbornsol ] , the disorder scattering time is found to satisfy the lower bound @xmath99 above which the ts state quasiparticle gap @xmath100 . in this paper , we will also assume that @xmath101 ( @xmath102 is the fermi energy with respect to the bottom of the top - most filled band ) need not be strictly enforced because of the existence of multiple bands in the actual experimental system , which will evade the localization problem . from eqs .
[ musm ] and [ taut ] , we find that the ts state gap induced on the fermi surface survives as long as the mobility of the system exceeds the threshold @xmath103 this suggests that the mobility threshold can be decreased by decreasing the kramer s pair overlap @xmath104 , which , as we show in eq .
[ krameroverlap ] , can be achieved by reducing the zeeman potential .
we will apply all these results to electron and hole - doped semiconductors in the following two sections .
in this section we review the two recently proposed systems , electron- and hole - doped semiconductors , to which we will apply our results derived in the previous section . for 2d electron - doped semiconductors with rashba and zeeman couplings , @xmath51 in eq .
[ hsm ] can be written as @xmath105 where @xmath20 is the effective mass of the electrons , @xmath11 is the rashba spin - orbit coupling constant , and @xmath9 is the zeeman splitting . for a one - dimensional nanowire along the @xmath106-axis , the above hamiltonian reduces to @xmath107 for a chemical potential @xmath108 , the above rashba spin - orbit - coupled hamiltonian has a single band at the fermi - level with an eigenstate described by the spinor @xmath109}}\nonumber\\ & \left(\begin{array}{c}\alpha k_f\\-\sqrt{v_z^2+\alpha^2k_f^2}-v_z\end{array}\right).\end{aligned}\ ] ] now , using eq .
[ delta ] , the superconducting quasiparticle gap in the clean electron - doped case is given by @xmath110 the proximity - induced pairing potential in the semiconductor , @xmath21 , can be related to the pairing potential in the bulk superconductor @xmath19 by the relation @xcite @xmath111 where @xmath112 here @xmath113 represents the tunneling matrix element between the semiconductor and the superconductor , @xmath114 is the normal state density of states of the superconductor and @xmath55 gives the rate at which electrons from the semiconductor tunnel into the superconductor .
@xcite we note that for weak - tunneling @xmath115 , and for strong tunneling , @xmath116 . the superconducting proximity effect in electron - doped inas / superconducting al interfaces
has been already been observed to lead to @xmath117 k , @xcite so that one can infer that the transparency parameter @xmath118 .
the kramer - pair overlap @xmath119 , which determines the effect of disorder scattering via eq .
[ scbornsol ] , is then given by @xmath120 which in the limit @xmath121 becomes @xmath122 .
recently , it has also been proposed @xcite that hole - doped nanowires , because of the possibly stronger spin - orbit coupling @xcite and larger effective mass of holes , might be a better candidate for creating topological superconductors and majorana fermions .
the hole - bands of a group iii - v semiconductor such as inas or gaas are composed of @xmath7-orbitals , which have orbital angular momentum @xmath123 .
because of strong spin - orbit coupling and cubic symmetry of these semiconductors , the description of the states near the top of the valence band at @xmath124 must be written in terms of spin-@xmath125 total angular momentum matrices @xmath126 .
the total spin-1/2 manifold of states forms a split - off band , which is separated by a large energy and therefore can be ignored .
the resulting three - dimensional hamiltonian is known as the four - band luttinger model @xcite , which is written as @xmath127\nonumber\\ & -\lambda_3\frac{1}{2 m_0}\sum_{\alpha\neq\beta = x , y , z}p_\alpha
p_\beta j_\alpha j_\beta,\end{aligned}\ ] ] where @xmath128 and @xmath129 are effective luttinger parameters .
the effective mass @xmath130 for the holes is negative so that the curvature of the valence bands are inverted with respect to the conduction bands .
we will consider a nanowire formed by confining the bulk hole - doped semiconductor confined strongly along the @xmath95 and @xmath53-directions , so that the confinement energy scale @xmath131 is large compared to all energy - scales we will be interested in .
therefore one can project the three - dimensional hamiltonian @xmath132 into the lowest confinement band of the nanowire by replacing @xmath133 , @xmath134 and adding a rashba spin - orbit coupling proportional to @xmath11 @xcite and a zeeman term proportional to @xmath9 to account for electric field induced inversion symmetry breaking and magnetic field - induced time - reversal symmetry breaking respectively .
the resulting hamiltonian for a hole - doped nanowire along @xmath106 with a rashba coupling ( for the lengths in the transverse directions @xmath135 ) is given by , @xcite @xmath136 where @xmath137 .
the hamiltonian at @xmath138 , commutes with @xmath139 .
moreover , the states with @xmath140 form the highest energy states in the valence band .
these states also have the lower effective mass of the two manifolds @xmath141 of states .
note that this is opposite to the two - dimensional case , where the heavy - hole is closer to the band - edge .
thus the light - hole states with @xmath142 form the set of states closest to the band - edge and we can restrict to this two - dimensional subspace @xmath142 , where we identify @xmath143 as a pseudo - spin variable . in this subspace , @xmath144 and @xmath145 , so that the hamiltonian for the two upper bands effectively becomes @xmath146 the hole hamiltonian eq .
[ rashbahole ] , in the specific isotropic limit we consider , is identical ( up to sign ) to the one - dimensional electron - doped rashba hamiltonian in eq .
[ rashba1d ] and therefore one can expect most of the physics discussed in the previous sub - section for the electron - doped nano - wire in this limit to translate to the hole - doped case , except for the different effective mass and rashba spin - orbit coupling . while the band - structure of the hole - doped wire is quite similar to the electron - doped case , it is not obvious from this discussion that the @xmath0-wave proximity effect has the same form in the @xmath49 spin - space . in principle , the proximity - induced effective superconducting pairing potential @xmath21 at an interface between the semiconductor and the superconductor can be calculated by projecting the anomalous self - energy @xmath147 induced by proximity to the superconductor , @xcite into the low - energy space of states @xmath148 .
since these states form a kramers pair , one can see that time - reversal symmetry only allows a conventional spin - singlet pairing in the @xmath149 space .
furthermore , the sub - space of states in the @xmath148 basis may be expanded in terms of the original spin eigenstates @xmath150 and @xmath151 ( which are related to the orbital angular momentum eigenstates @xmath152 by a unitary transformation ) as @xmath153\\ & { \left\vert { \sigma_y=-\frac{1}{2 } } \right\rangle}\equiv { \left\vert { j_y=-\frac{1}{2 } } \right\rangle}\nonumber\\ & = \frac{1}{\sqrt{3}}[{\left\vert { l_x=0,s_y=\frac{1}{2 } } \right\rangle}-i{\left\vert { l_z=0,s_y=\frac{1}{2 } } \right\rangle}\nonumber\\ & -{\left\vert { l_y=0,s_y=-\frac{1}{2 } } \right\rangle}].\end{aligned}\ ] ] to calculate @xmath21 , which is related to the matrix - element @xmath154 of a spatially smoothly varying ( on the orbital scale ) proximity - induced self - energy @xmath147 , one can ignore the matrix elements involving the orbital @xmath155 for sc interfaces normal to the @xmath156-axis .
this leads to the effective barrier transparency parameter @xmath55 in eq .
[ delta0 ] to be reduced by a factor of @xmath157 .
the corresponding pairing potential @xmath21 from eq .
[ delta0 ] is not significantly affected and remains the same between electrons and holes provided a moderately transparent barrier with @xmath158 can be achieved .
therefore , more generally , one can expect a proximity - effect where the transparency - factor of the sm / sc interface , @xmath55 , is suppressed by at most a factor of order 1 , which a value for the pairing potential @xmath159 , even for holes . to summarize , while the hole - doped nanowires have fundamentally different microscopic hamiltonians consisting of a triplet of orbitals with @xmath7-symmetry , for nanowires it is possible to work in the limit of symmetric confinement along the two directions transverse to the wire , where we have shown that the effective hamiltonian and proximity - effect are identical in form to the electron - doped wire case .
of course , the parameters in the hamiltonian are expected to be quite different and in particular the effective mass and rashba spin - orbit coupling are expected to be much larger .
@xcite at the same time , the corresponding mobility for the semiconductors that can be achieved are found to be significantly lower , @xcite indicating much stronger disorder for hole - doped systems .
thus , the hole systems have the advantage ( disadvantage ) of stronger so coupling ( larger disorder ) , and whether the net prospect for realizing the non - abelian majorana mode is greater in hole - based systems ( than the electron - based systems ) or not will depend entirely on which of these two effects wins over in realistic samples .
another possible advantage of the hole systems is the higher value of the lande g - factor which necessitates weaker magnetic field values in achieving the condition for topological superconductivity .
one should , however , realize that very little actual experimental work exists in hole - based semiconductor nanowires .
however , it is encouraging that the superconducting proximity effects in both electron and hole - based semiconductors have been observed in experiments .
@xcite therefore , we do believe , that the hole - based systems ( e.g. inas , insb , sige , and perhaps even gaas ) should be included in the experimental search for the majorana mode in semiconductor - superconductor heterostructures .
the clean - limit quasiparticle gap @xmath3 in the topological superconducting state can be obtained for the rashba spin - orbit coupled hamiltonian eq .
[ rashba ] using eqs .
[ delta1 ] and [ delta0 ] as @xmath160 this expression suggests that to obtain a large gap , @xmath3 , in the semiconductor one should restrict oneself to the limit where @xmath161 .
moreover , to remain in the requisite odd sub - band ( confinement bands ) limit it is also necessary to choose a chemical potential @xmath28 in the gap ( at @xmath124 ) induced by the zeeman splitting @xmath9.@xcite this restricts @xmath162 in the highest filled sub - band . in this limit
one can approximate @xmath163 , which we call the spin - orbit energy @xmath164 .
this spin - orbit energy @xmath164 is a crucial energy scale in the problem which is to be compared with the other independent energy scales such as @xmath9 , @xmath165 and @xmath24 .
therefore , from eq . 30 , to obtain a large quasiparticle gap at the fermi surface
the system must be restricted to the regime @xmath166 the semiconductor system is in the topological state if and only if the pfaffian invariant of the relevant bdg hamiltonian is negative , @xcite which is equivalent to the topological condition @xmath167 where @xmath168 , @xmath169 and @xmath170 are the renormalized parameters in the effective hamiltonian of the proximity coupled semiconductor after the superconductor has been integrated out .
@xcite thus , using eq .
[ delta0 ] and eq . [ topcons1 ] , the topological condition in terms of the bare parameters can be written as , @xmath171 eq
. [ topcons ] and eq . [ vzcons ] , together with eq .
[ scbornsol ] ( or eq . [ mut1 ] ) , define the optimal constraints on the material considerations for the realization of the ts phase in the semiconductor .
in what follows , we assume that the system is reasonably deep in the topological superconducting phase ( i.e. @xmath172 ) so the smallest gap occurs near the fermi wave - vector @xmath173 instead of near @xmath174.@xcite therefore in attempting to optimize the quasiparticle gap one notices from eqs .
[ vzcons ] , [ topcons ] a hierarchy of energy scales @xmath175 the first two - inequalities from the left are constraints between the external parameters @xmath176 and @xmath55 , out of which @xmath9 and @xmath55 must be tuned to satisfy the above inequality . the other external parameter is the disorder scattering rate in the semiconductor , @xmath177 , which must be reduced to satisfy @xmath178 . the two - inequalities on the right - hand side of eq .
[ ineq ] are between the derived quantities @xmath3 and @xmath94 , which are completely determined once @xmath179 are given , so that the condition @xmath180 and @xmath181 follow as direct consequences of eqs .
[ delta5 ] and eq .
[ scbornsol ] , respectively .
therefore , the disordered ts state quasiparticle gap , @xmath94 , is constrained by the spin - orbit energy scale @xmath164 , which must be as large as possible in the problem . in the presence of disorder
, one can use eq .
[ scbornsol ] together with eq .
[ krameroverlap ] to show that for @xmath182 , the disorder renormalized ts state quasiparticle gap is given by @xmath183 where @xmath184 .
the mobility threshold , which is the minimum mobility required to realize a non - zero ts state gap @xmath94 in the presence of disorder , is obtained by using eqs .
[ mut1 ] , [ krameroverlap ] to be @xmath185 which vanishes in the limit @xmath186 .
note we have reintroduced the @xmath187 to make the expression dimensionally consistent . since @xmath188 increases with @xmath3 , which in turn increases with @xmath55 , for deriving eq .
[ mut2 ] , we have assumed the @xmath55 is chosen to reach its maximum value @xmath189 , above which the system can no longer be considered to be deep in the topological phase ( see eq .
[ ineq ] ) .
in fact , we will assume that @xmath9 has been tuned to be @xmath190 for the rest of the paper . as a function of semiconductor mobility @xmath188 for electron - doped wires with @xmath191 , @xcite and @xmath192 allowing for future electric - field enhanced spin - orbit coupling .
@xcite ( b ) the values of @xmath193 realizing @xmath94 in panel ( a ) plotted with @xmath188 .
the gaps in the adjacent bulk superconductor are taken to be @xmath194 corresponding to al and @xmath195 corresponding to nb . ]
of course , from eq .
[ topcons ] , there is a threshold value for @xmath9 only above which the ts state itself is realized .
therefore , @xmath9 can not be taken as zero in eq .
[ qpgap ] to maximize the disordered ts state gap @xmath94 . in eq .
[ qpgap ] , to deduce the optimum @xmath94 , we use the r.h.s of eq .
[ delta5 ] to substitute for @xmath3 and , for a robust topological state , we take @xmath189 .
the r.h.s of eq .
[ qpgap ] is then a function of @xmath9 , @xmath78 ( or @xmath196 ) , @xmath19 and @xmath164 . among these @xmath19 ( the gap in the adjacent superconductor ) and @xmath164 ( the spin - orbit coupling energy in the semiconductor )
are known experimental quantities .
therefore eq .
[ qpgap ] defines a formula for @xmath94 in terms of @xmath9 parametrically related by @xmath188 . for given values of @xmath188
, we maximize @xmath94 with respect to @xmath9 and plot these maximized values of @xmath94 as a function of @xmath188 for electron and hole - doped semiconductors in figs .
[ fig1 ] and [ fig2 ] , respectively .
these plots are the central results of this paper . as a function of semiconductor mobility @xmath188 for hole - doped wires with @xmath197 .
@xcite ( b ) the values of @xmath193 realizing @xmath94 in panel ( a ) plotted with @xmath188 .
the gaps in the adjacent bulk superconductor are taken to be @xmath194 corresponding to al and @xmath195 corresponding to nb . ] from fig .
[ fig1 ] it is clear that for electron - doped wires with a mobility of @xmath198 @xmath14/v - s @xcite , which has been achieved for wires with a diameter @xmath199 nm , a measurable robust ts state gap of @xmath15 mk is achievable .
as is also clear from fig .
[ fig2 ] , for hole - doped wires with a significantly larger spin - orbit energy @xmath16 mev , a larger gap of @xmath17 k is achievable for a mobility of @xmath200 @xmath14/v - s . @xcite it should be noted that in general the holes have much lower mobilities than electrons in the same semiconductor material even after taking into account their large effective mass difference .
in this paper we have addressed the effects of disorder from the semiconductor surface on the stability of the topological superconducting state on 1d and 2d electron- and hole - doped semiconductors proximity coupled to bulk 3d @xmath0-wave superconductors . in recent works
@xcite it has been shown that the effects of disorder from the adjacent bulk superconductor on the ts state is minimal , so we have ignored this effect . however , since the ts state in the semiconductor explicitly breaks the time - reversal symmetry ( due to an external zeeman splitting ) , even the disorder from the semiconductor surface itself has an effect on the stability of the ts state and will close the bcs - like superconducting quasiparticle gap at the semiconductor fermi surface for low enough mobility .
we have shown that the disordered ts state quasiparticle gap @xmath94 is suppressed with increasing values of the ratio @xmath12 between the zeeman splitting @xmath9 and the spin - orbit energy scale @xmath201 , @xmath202 . the dimensionless quantity @xmath12 measures the extent of the tr symmetry breaking in the semiconductor to produce the ts state and majorana fermions .
our main results are plotted in figs .
[ fig1 ] and [ fig2 ] .
they display the achievable disorder renormalized ts state gap @xmath94 as a function of the semiconductor mobility @xmath188 in both electron and hole - doped semiconductor wires . from fig .
[ fig1 ] we find that for electron - doped wires with a mobility of @xmath198 @xmath14/v - s @xcite , which has been achieved for wires with a diameter @xmath199 nm , a measurable robust ts state gap of @xmath15 mk is achievable . from fig .
[ fig2 ] we find that for hole - doped wires with a significantly larger spin - orbit energy @xmath16 mev , a larger gap of @xmath17 k is achievable for a mobility of @xmath200 @xmath14/v - s .
@xcite thus , despite the fact that the semiconductor ts state breaks the time - reversal symmetry , realizing a robust ts state is possible in both electron- and hole - doped systems even in the presence of realistic disorder on the surface of the semiconductor . before concluding
, it may be useful to provide a qualitative discussion and a general perspective on the issue of optimal materials for the realization of the predicted ts phase and majorana modes in semiconductor - superconductor sandwich structures .
the first point to emphasize is that there are far too many independent physical parameters entering the problem for theory to be particularly useful in this context .
at least six independent material parameters play a role here : @xmath11 , @xmath22 , @xmath9 , @xmath28 , @xmath19 , @xmath113 ( which determines @xmath55 in eq .
[ lambda ] ) .
in addition , disorder , which plays a crucial role by the non - existence of anderson s theorem in the time - reversal invariance broken ( i.e. @xmath203 ) situation , is itself characterized by several independent physical mechanisms : disorder in the adjacent bulk superconductor , long - ranged disorder in the semiconductor , disorder - induced fluctuations in @xmath113 , interface disorder , puddles , and so on . a physical system characterized by such a large ( @xmath204 10 ) set of independent parameters , all of which are important in determining the fate of the ts state , is not an easy optimization problem to deal with theoretically . in particular
, it may turn out to be more practical to simply carry out experiments on available systems ( e.g. inas or insb on al or nb ) than to trust precise theoretical quantitative predictions with respect to the optimal materials choice . with these important caveats in mind , we can
, nonetheless , make some general qualitative statements , quite apart from the more precise quantitative statements made in the main body of this work . in the clean case , where disorder is not an issue ( i.e. extreme high - mobility semiconductor systems ) , it is obvious that the semiconductor spin - orbit coupling should be as large as possible .
this continues to be the case even in the presence of disorder , making @xmath205 to be the key physical parameter , which should be made as large as possible ( including enhancement by external gates if feasible @xcite ) with no constraints .
similarly , disorder is always harmful to the ts state , and therefore @xmath24 should be made as small as possible by working with the highest mobility materials .
although having a spin - orbit coupling strength as large as possible is obviously helpful to the realization of the ts phase , this may have the detrimental effect of enhancing impurity - induced scattering between the spin - split bands , thus reducing the mobility , @xcite and therefore some optimization may be necessary even with respect to spin - orbit coupling and mobility in order to satisfy the general requirement of having both the highest possible spin - orbit coupling and mobility .
the same is to a limited degree true of the superconducting gap @xmath19 ( or equivalently @xmath206 ) in the bulk supercnoductor , but this advantage is considerably nullified by the fact that the proximity effect is determined primarily by the parameter @xmath55 ( see eq .
[ delta0 ] ) , and this parameter should be made as large as possible , even at the cost of having a lower @xmath19 in the bulk superconductor ( e.g. al may be preferable over nb if @xmath207 ) .
we note that having an arbitrarily large proximity gap @xmath21 in the semiconductor may not necessarily help unless @xmath164 can also be simultaneously increased ( e.g. by an external gate ) since @xmath208 is a necessary condition for the existence of the ts phase .
similarly , having a very small @xmath9 does not help since @xmath209 determines the condition for a robust topological phase , and having a small @xmath9 necessarily requires having a much smaller @xmath3 , which severely restricts the temperature range where the experiments investigating the topological phase can occur since @xmath210 is necessary .
for the same reason a small @xmath164 , e.g. gaas , with correspondingly smaller @xmath9 and @xmath3 would not work either since the operational temperature would then be too small .
as for disorder , our discussion in this work has considered the semiconductor disorder to be short - ranged , which is consistent with interface disorder ( or roughness ) scattering which is known to be important in semiconductor nanowires .
carrier screening should render long - range coulomb disorder arising from random charged impurities in the semiconductor into effective short - ranged disorder , making our model of rather general validity .
the fact that we model disorder through an effective relaxation time @xmath78 defined through the measured carrier mobility makes the model robust with respect to different types of disorder in the semiconductor , but we note that for long - ranged coulomb disorder @xmath177 is really only a lower bound on the effective disorder .
( disorder arising from impurities in the superconductor does not affect the ts phase at all , see ref .
@xcite for details . ) a simple dimensional argument , which has been authenticated by a renormalization group calculation , @xcite suggests @xmath211 , i.e. @xmath212 , as the bound on the effective scattering rate for the topological gap @xmath3 to survive disorder in a spinless chiral p - wave superconductor characterized only by these two parameters ( unlike our problem which is characterized by many physical parameters ) .
we note that our microscopic born approximation calculation ( eq . [ taut ] ) is consistent with this case deep in the ts phase for @xmath213 . for @xmath214 , the effect of disorder scattering
is suppressed so that the topological gap survives disorder for a broader range of parameters @xmath215 .
unfortunately for the semiconductor - superconductor sandwich structures of interest to us , @xmath215 only provides a necessary condition since @xmath3 can not be made arbitrarily large unless @xmath164 i.e. the spin - orbit coupling is itself arbitrarily large .
the sufficient conditions for the gapped topological state to exist in the semiconductor sandwich structures are given by eqs .
33 , 34 , which require large " @xmath9 making it necessary to have even larger @xmath164 !
thus , the existence of the ts phase in the realistic semiconductor sandwich structures depends on the seamless and multiparameter materials optimization of spin - orbit coupling , spin splitting , superconducting gap , and disorder in an inter - dependent manner .
we conclude with the observation that the single most important ingredient for the ts state to emerge in semiconductor - superconductor sandwich structures is to have a large spin - orbit coupling and high mobility . for electrons both inas and insb satisfy the necessary requirements if the mobility can be enhanced around @xmath216 @xmath14/v - s . for holes the spin orbit coupling being very large , much lower mobilities ( @xmath217 @xmath14/v - s ) should work .
these considerations imply that holes should be looked into experimentally for the realization of the ts phase .
in addition , having several ( odd number of ) subbands @xcite occupied in the nanowire should help since it reduces disorder through screening and enhances the effective values of @xmath164 .
we emphasize that the existence of the ts state depends delicately on the competition among superconducting proximity effect , spin - orbit coupling , spin splitting , and chemical potential .
( disorder scattering always suppresses the ts state . ) as such , any materials optimization must carefully choose among these physical parameters while at the same time maintaining the highest possible mobility ( i.e. smallest possible disorder scattering ) .
the experimental search for the majorana mode should thus concentrate on semiconductor systems of the highest mobility and the strongest so coupling using the appropriate superconductors which produce the best quality interface between the superconductor and the semiconductor .
this work is supported by darpa - mto , darpa - quest , jqi - nsf - pfc , microsoft - q , and nsf .
we thank chuanwei zhang for helpful discussions .
in this appendix , we start by first simplifying the dyson equations eqs .
[ dyson ] and [ sigma ] and then solve the equations for the first order born approximation case . the self - consistent born approximation is solved in appendix .
[ selfconsistentborn ] substituting eqs .
[ vmat ] and [ disordervertex ] into eq .
[ sigma ] , the self - energy @xmath79 is written as @xmath218.\label{sigmad}\end{aligned}\ ] ] in principle , the disorder averaged scattering potential @xmath84 involves integrals of 4 bloch eigenstates @xmath219 of the form @xmath220 where the sum over @xmath49 implies that the disorder is spatially local spin - independent disorder .
having several sub - bands occupied in the 1d nanowire , as has recently been proposed @xcite , should help the situation considerably , both by expanding the regime of stability of the ts phase and by screening out the disorder due to higher carrier density . substituting the wave - function into the matrix elements ,
we get @xmath221 substituting the matrix elements into the disorder - induced self - energy we find that , @xmath222.\label{sigma1}\end{aligned}\ ] ] to solve dyson s equations , it is convenient to define the functions @xmath223 by the relation @xmath224 substituting this into the dyson equations ( eq . [ dyson ] ) together with the expression for the self - energy ( eq . [ sigma1 ] ) in the lowest order born approximation we get @xmath225 where @xmath226 , @xmath227 .
the integrals in eq .
[ borneq ] can be written analytically as @xmath228e^{i\theta_{\bm k}}\end{aligned}\ ] ] where @xmath229 and @xmath230 remains unrenormalized .
the lowest - frequency pole in the disorder averaged green function @xmath231 given by eq .
[ ginv ] , which corresponds to the quasiparticle gap , occurs when @xmath232 and @xmath233 .
the disorder averaged quasiparticle gap @xmath86 can then be obtained by solving @xmath234=\delta[1+\frac{f_1}{\sqrt{\delta^2-\omega^2}}].\ ] ] for sufficiently weak disorder @xmath235 , the quasiparticle gap @xmath236 , so that we can expand in the reduction of the quasiparticle gap @xmath237 to lowest order and obtain the quasiparticle gap reduction to be @xmath238 this leads to the pole of the disordered average green function @xmath87 which is also the ts state quasiparticle gap in eq .
[ eq : x ] @xmath239.\ ] ]
in this appendix we provide an exact solution for the quasiparticle gap within the self - consistent born approximation . to do this , we start with the self - consistent version of eq .
[ borneq ] , @xmath240 performing the @xmath241 integral in the second equation we immediately obtain @xmath242 . performing the @xmath241 integral in the other 2 equations @xmath243 where the square root is chosen so that the real part of @xmath244 , @xmath245 and @xmath246 . writing @xmath247 ,
the pair of equations can be combined to a single equation for @xmath106 @xmath248 at @xmath249 , the equation has 4 real roots @xmath250 and @xmath251 .
the quasiparticle gap is defined by the value of @xmath86 above which the green function @xmath252 has a finite imaginary part .
this corresponds to the value of @xmath86 where 2 of the real roots of @xmath106 merge into a pair of complex roots , so that eq .
[ eqy ] has 2 real and 2 complex roots . using the properties of quartic polynomial equations , the transition point between 4 real roots and 2 real roots , is signalled by the vanishing of the discriminant of the quartic polynomial .
solving for the frequency @xmath86 where the discriminant vanishes leads to the solution @xmath253 where @xmath254 . in the 2d rashba and 1d superconducting case
, we get the solution of the ts state quasiparticle gap in eq .
[ scbornsol ] @xmath255^{1/2}.\end{aligned}\ ] ] | in proximity to an @xmath0-wave superconductor , a one- or two - dimensional , electron- or hole - doped semiconductor with a sizable spin - orbit coupling and a zeeman splitting can support a topological superconducting ( ts ) state .
the semiconductor ts state has majorana fermions as localized zero - energy excitations at order parameter defects such as vortices and sample edges . here
we examine the effects of quenched disorder from the semiconductor surface on the stability of the ts state in both electron- and hole - doped semiconductors . by considering the interplay of broken time reversal symmetry ( due to zeeman splitting ) and disorder
we derive an expression for the disorder suppression of the superconducting quasiparticle gap in the ts state .
we conclude that the effects of disorder can be minimized by increasing the ratio of the spin - orbit energy with the zeeman splitting . by giving explicit numbers
we show that a stable ts state is possible in both electron- and hole - doped semiconductors for experimentally realistic values of parameters .
we discuss possible suitable semiconductor materials which should be the leading candidates for the majorana search in solid state systems . |
glasses are structurally disordered systems . as common experience shows , in the macroscopic limit they support sound waves as the corresponding crystalline materials do .
in fact , averaging on a length - scale large enough , the details of the microscopic arrangement become essentially irrelevant .
this holds true for sound waves with long wavelengths of at least several hundreds of nanometers , corresponding to wavenumbers , @xmath1 , in the 10@xmath2 nm@xmath3 range , as those probed with light scattering techniques @xcite . on further increasing @xmath1 ,
the effect of the structural disorder must at one point appear .
the much larger @xmath1-scale of few nm@xmath3 is also well known , since it can be accessed experimentally with inelastic x - ray @xcite and neutron @xcite scattering techniques , and numerically with molecular dynamics simulations @xcite .
these studies typically probe the dynamic structure factor , @xmath4 , that is the space and time fourier transform of the density - density correlation function .
they clearly indicate the existence in glasses of excitations that appear in those spectra as very broad peaks whose position as a function of @xmath1 is characterized by a sinusoidal - like dispersion curve .
thus , these excitations strongly recall the acoustic modes in ( poly-)crystalline systems up to roughly one half of the pseudo brillouin zone @xcite , i.e. down to distances corresponding to the interatomic spacing . for this reason ,
they are often dubbed as acoustic - like . however , as their broadening clearly indicates , they are far from being crystal - like modes and correspond in fact to a complex pattern of atomic motions @xcite .
unfortunately , experimental and numerical studies leave a gap between the few nanometers scale and the hundreds of nanometers one that is extremely difficult to access .
this keeps still open a number of fundamental questions in the physics of glasses and , more in general , on the nature of the vibrational excitations in disordered systems . _
i ) _ how does the transition for the acoustic - like excitations look like between the small-@xmath1 debye - like behaviour and the large-@xmath1 regime ?
in other words , how does it happen that reasonably well defined plane waves transform , on increasing @xmath1 , into a complex pattern of atomic motions that mirror the structural disorder ?
_ ii ) _ similarly to crystalline systems , glasses are characterized by a universal behaviour in some fundamental low - temperature observables like specific heat and thermal conductivity @xcite . in particular ,
in the @xmath010 k range the specific heat shows an excess with respect to the @xmath5 , debye - model prediction and the thermal conductivity shows a plateau . despite of a longstanding debate @xcite , their origin is still poorly understood .
it is however generally accepted that they are related to the ubiquitous existence at frequencies of @xmath01 thz of an excess of modes in the vibrational density of states , @xmath6 , over the debye - model prediction @xmath7 , with @xmath8 the debye frequency .
this excess of modes is best visible in the reduced density of states , @xmath9 , where it appears as a broad feature known as boson peak @xcite . is then this universal behaviour in the low - temperature thermal properties or in the vibrational density of states related to the peculiar nature of the acoustic - like excitations in the low - frequency range ? or , alternatively , do we have to imagine that glasses are characterized by additional low - frequency modes ? _
iii ) _ what is the information content that we can extract from the high-@xmath1 acoustic - like excitations measured in inelastic scattering experiments or calculated numerically ?
in which physical properties are their crystal - like features ( e.g. existence of dispersion curves ) reflected ?
the experiments that have attempted to answer these questions accessing the difficult 10@xmath21 nm@xmath3 @xmath1-range lead to contrasting interpretations .
an experiment based on a tunnel junction technique reported @xcite linear dispersion for the transverse acoustic excitations in a silica glass up to frequencies 50% smaller than the boson peak position in that glass ( @xmath01 thz ) .
this result indicated that the acoustic excitations are unaffected in the frequency range relevant for the thermal properties in the @xmath010 k range , and seemed to exclude any acoustic contribution to the low - temperature anomalies in the specific heat of glasses @xcite .
early inelastic x - ray scattering results @xcite seemed to confirm that scenario and showed crystal - like dispersing high - frequency longitudinal acoustic - like excitations with a broadening increasing quadratically with the frequency .
again , a simple linear dispersion of the longitudinal acoustic - like modes was observed at frequencies corresponding to the boson peak position , thus suggesting a smooth transition between the macroscopic and microscopic regime .
recent and more accurate inelastic x - ray scattering studies have however revealed that the boson peak marks the energy where a qualitative change takes place : the longitudinal acoustic - like excitations show , below the boson peak position , a marked decrease of the phase velocity @xcite and a broadening characterized by a remarkable fourth - power - law frequency dependence @xcite .
a similar behaviour for the broadening of the transverse acoustic modes had been found in a silica glass at low temperature and at frequencies below the boson peak position using a tunnel junction technique @xcite .
conversely , a recent experiment using inelastic ultraviolet light scattering to measure the longitudinal acoustic modes in a silica glass at room temperature reported the onset of this peculiar @xmath10 regime at frequencies one order of magnitude smaller than the boson peak position @xcite : the boson peak would then be not directly related to this regime .
a complex and sometimes contradictory picture seems therefore to come out of the experiments performed so far .
classical molecular dynamics simulations have provided a complementary tool to study the vibrational properties of glasses , starting from the pioneering investigations of rahman and coworkers @xcite . the body of results available until now supports a scenario where the longitudinal acoustic - like excitations seem to be largely decoupled from the boson peak : they show a linear dispersion and a broadening that grows quadratically with frequency with no special feature in the frequency region where the boson peak appears @xcite .
however , these studies could not provide a final answer on this issue due to the fact that the largest wavelengths that could be studied so far fixed by the simulation box size and then ultimately by computer power were still in the range of few tens of interatomic spacings . as a consequence ,
the corresponding lowest frequency acoustic - like excitations that were accessible lied too close to the boson peak position to allow for definite conclusions in this supposingly crucial frequency range .
we report here on a study of the vibrational dynamics of the classical lennard - jones ( lj ) monatomic glass model @xcite .
the reason of our choice is simple : despite of the fact that this system easily crystallizes below the melting temperature and thus requires extremely fast and experimentally out of reach quenching rates in order to prepare a glass starting from the melt , it is the simplest realization of a structural glass at our disposal .
it thus provides the very basic ingredients of the vibrational dynamics of a structural glass that in other systems might be superimposed to additional more complex effects .
specifically , we present here results for an exceptionally large simulation box containing up to @xmath11=10@xmath12 particles and clarify how the acoustic modes look like in the frequency region where the boson peak appears . in particular
, we managed to probe acoustic excitations down to frequencies one order of magnitude lower than the boson peak position .
thanks to that , the study of their @xmath1-dependence allows us to establish that the boson peak originates from a deformation of the dispersion curves with respect to the crystal case .
this boils down to a direct connection between the boson peak and the breakdown of the debye continuum approximation for the acoustic excitations that takes place on a length - scale matching that of the medium range order of the glass @xcite .
details about computer simulation methods used are given in appendix [ appendix - methods ] .
we have calculated both the transverse , @xmath13 , and the longitudinal , @xmath14 , dynamic structure factors which give information on the transverse and longitudinal acoustic - like excitations , respectively .
it is important to underline that while the latter can be obtained experimentally using scattering techniques , the former can only be studied in the frequency range relevant here using computer simulations . in
what follows we will use lj units . to make connection with experiments
, we recall that , if we model argon using the lj potential , then the temperature scale is in units of @xmath15=125.2 k , the length scale in units of @xmath16=3.405 @xmath17 , and the time scale in units of @xmath18=2.11 ps . in fig . 1 some representative @xmath13 spectra
are reported , including those corresponding to the lowest @xmath1 value that we could reach in our simulation .
, for a lj glass at number density @xmath19=1.015 , temperature @xmath20=10@xmath21 and at the indicated @xmath1 values , including the smallest one accessible using a simulation box containing @xmath010@xmath12 atoms
. the brillouin peaks shift towards higher frequencies and show a clear broadening on increasing @xmath1 .
, scaledwidth=48.0% ] for reference , we recall that at the studied number density @xmath221.015 ( v being the simulation box volume ) the melting temperature of the lj system is @xmath23 1 , and the glass - transition temperature @xmath24 0.4 @xcite . for what concerns the glass that we study , the debye frequency and wavenumber are @xmath8=16.2 and @xmath25=3.92 , respectively ; the first sharp diffraction peak is at @xmath26 7 , that corresponds to an average nearest neighbours distance of @xmath270.9 .
the spectra reported in fig . 1
refer then to @xmath1-values down to @xmath010@xmath28 times smaller than the border of the pseudo - brillouin zone located at @xmath29 .
the @xmath13 spectra are characterized by two symmetric peaks ( brillouin peaks ) in addition to a sharp elastic peak at @xmath300 .
the brillouin peaks can be characterized by the position of the maximum and the broadening .
this information has been obtained by fitting a damped harmonic oscillator model to the spectral region , @xmath31 , around the brillouin peaks : @xmath32 the parameters @xmath33 and @xmath34 represent the characteristic frequency and broadening ( fwhm ) of the brillouin peaks , respectively .
the parameter @xmath35 is utilized to obtain the transverse sound phase velocity , @xmath36 , and is reported in fig .
2a as a function of frequency . at high - frequencies and @xmath37 at low - frequencies .
the transition between the two regimes appears at the frequency where the phase velocity in a ) shows a minimum .
, scaledwidth=48.0% ] these data show an increase with frequency ( positive dispersion ) of the sound velocity for @xmath38 as already reported for the longitudinal excitations in the same glass @xcite .
it is interesting to observe that below that frequency the macroscopic @xmath300 sound velocity limit is not directly recovered : instead a previously unnoticed region where the phase velocity decreases with increasing frequency ( softening ) appears .
this is exactly the region where the boson peak is found in this lj glass ( see fig .
4b below ) .
the boson peak then appears not in a frequency range where the acoustic - like excitations disperse linearly ( constant phase velocity ) , at odds with what was previously thought @xcite , but where they experience a more complex dispersion behaviour .
this is important because it clarifies that the debye continuum approximation for the acoustic excitations breaks down at frequencies comparable to the boson peak position , i.e. at frequencies much lower than previously expected .
complementary information on this issue can be found in fig . 2b , where the frequency - dependence of the broadening , @xmath39 , of the transverse acoustic - like excitations is presented .
these data clearly show two different regimes : a @xmath40 regime at high - frequencies , as already known from previous studies and associated to the structural disorder of the glass @xcite , and a previously unnoticed @xmath10 regime at low frequencies .
the frequency that marks the change of regime is very close to , though slightly lower than , the boson peak position ( see fig .
moreover , the low frequency regime appears in the same range where the softening of the sound velocity shows up in fig .
2a , thus indicating that these two features must be connected .
this , however , is not surprising if one recalls that the brillouin position and broadening can be related to the real and imaginary part of a complex self - energy , respectively @xcite .
= 59 . within error - bars
, we can conclude that the frequency - dependence of the longitudinal phase velocity directly reflects that of the transverse one . in panel
b ) the longitudinal excitations broadening data are compared to the corresponding transverse ones ( red circles , right axis ) : the two sets of data can be convincingly scaled one on top of the other . within error - bars , we can again conclude that the frequency - dependence of the longitudinal data directly reflects that of the transverse ones .
, scaledwidth=48.0% ] a similar scenario holds for the longitudinal acoustic - like modes .
the longitudinal sound velocity data ( blue squares ) are reported in fig .
3a , and show again a decrease at low frequency followed by a positive dispersion for @xmath38 .
the similarity with the behaviour of the transverse data suggests a common origin for both .
indeed , the longitudinal and transverse sound velocities in an isotropic elastic medium are simply related by the expression : @xmath41 where @xmath42 is the mass density and @xmath43 is the bulk modulus .
this equation simply tells that in a glass the shear modulus , @xmath44 , the longitudinal modulus , @xmath45 , and the bulk modulus @xmath43 are related , and only two of them give independent information . the low - frequency
, macroscopic value for the bulk modulus @xmath4659 can be obtained from the low - frequency data for @xmath47 and @xmath48 , and is in good agreement with a literature value obtained for a slightly different system @xcite . assuming a frequency independent value for @xmath43
, we can then estimate the longitudinal sound velocity using the transverse data reported in fig .
the results of this calculation are shown in fig .
3a ( red circles ) .
the good correspondence between the two sets of longitudinal velocity data derived directly from @xmath14 or from @xmath13 via eq . [ eq2 ] strongly supports a scenario where : i ) the bulk modulus is frequency independent and ii ) the frequency dependence of the longitudinal sound velocity simply derives from that of the transverse one .
this picture is further reinforced by the comparison shown in fig .
3b between the broadening of the longitudinal acoustic - like excitations ( blue squares , left axis ) and that of the transverse ones ( red circles , right axis ) .
it is clear here that these two quantities , within error - bars , can be scaled one on top of the other .
this confirms that , within the accuracy of our calculation , the frequency dependence that characterizes the longitudinal acoustic - like excitations comes into play through the shear component of the longitudinal response , while the bulk component is a mere spectator .
the observation of the @xmath10 regime at low frequency in the acoustic attenuation for both polarizations is an interesting result and is predicted by several models @xcite .
the simplest one can be formulated in terms of rayleigh scattering of the acoustic excitations from some kind of structural disorder @xcite . in order to make this argument
quantitative one needs to identify the scattering units that produce it , a difficult task that has lead to different conclusions @xcite .
other approaches include a description of the disorder in terms of a continuum model with randomly fluctuating transverse elastic constants @xcite , or of lattice models with disorder in the elastic constants @xcite . on this basis
, it has been quite puzzling not to observe this strong scattering regime in previous molecular dynamics studies @xcite .
it has for example been argued that , in contrast with lattice based models , this feature would be absent in realistic glasses due to their intrinsic high level of frustration ( large internal stress ) @xcite .
the present results solve the issue , clarifying that a strong scattering regime is present in fact in realistic structural glasses also , in agreement with a number of theoretical @xcite and experimental @xcite results .
we believe that the reason why this regime got unnoticed before has to do with the fact that , beside the obvious cases of too small simulation boxes , the attention was usually focused on the longitudinal excitations ( the only ones experimentally accessible ) where this regime is difficult to study in detail even with a box as large as the one used here , see fig . 3b .
the present results in fact clarify that the signature of strong scattering appears for both polarizations at the same frequency which implies that as the longitudinal speed of sound is larger than the transverse one ( by a factor 2.3 in the present case ) the strong scattering regime appears for the longitudinal polarization at lower @xmath1 s than for the transverse one : its observation is then definitively less favorable for the longitudinal polarization .
-dependence of the lowest frequency transverse eigenvalues of the studied lj glass ( circles ) together with the function used to empirically describe it ( full line ) .
the initial slope of the curve corresponds to the @xmath490 limit of the transverse phase velocity of fig .
2a ( dashed line ) .
note that here @xmath1 is only an approximate quantum number , and that it is increasingly difficult to associate higher frequency eigenvalues to specific @xmath1 values .
b ) reduced density of states of the studied lj glass ( circles ) showing the boson peak at @xmath50 2 . the macroscopic , debye limit ( dashed line )
is indicated as well . the reduced density of states directly obtained from the data in fig .
4a via eq .
[ eq3 ] is reported ( full line ) up to the maximum frequency where this analysis is appropriate .
the agreement between the two calculations is good note that there is no adjustable parameter in this comparison .
this implies that the density of states , up to the boson peak , is well described by the low-@xmath1 inflection observed in fig .
, scaledwidth=48.0% ] more in general , the present results clearly show that the macroscopic and microscopic regimes for the acoustic - like excitations are connected by a crossover region where the acoustic - like dispersion curves show a considerable deformation with respect to a simple linear dependence : the sound velocities decrease with increasing frequency thus directly testifying the existence of an abrupt breakdown of the debye approximation in glasses @xcite .
this comes together with the signature of strong scattering for the acoustic - like excitations across which reasonably well defined plane waves transform into a complex pattern of atomic vibrations .
the connection to the boson peak is also clear : a softening of the transverse and longitudinal sound velocities implies an excess in the reduced vibrational density of states above the debye level . since this softening directly appears in the range where the boson peak is observed , or at the corresponding temperatures where the excess in the specific heat universally appears in glasses @xcite
, we can conclude that these anomalies must have an acoustic contribution . in the following
we make this connection quantitative . in order to gain further insight into these molecular dynamics results
, we performed a standard normal modes analysis of the glass in its inherent structures , i.e. we have diagonalized the dynamical matrix calculated in local minima of the potential energy landscape and derived eigenvalues and eigenstates .
the lowest frequency eigenstates of transverse polarization are close to be plane waves : it is possible to associate them with a leading @xmath1 value @xcite ( see appendix [ appendix - methods ] for details ) . in fig
4a we report the corresponding eigenvalues as a function of @xmath1 ( circles ) ; the reduced vibrational density of states derived from the complete normal modes analysis is shown in fig .
4b ( circles ) .
4a corresponds to a pseudo - dispersion curve : it is not formally a dispersion curve but only an approximate one since in a disordered system @xmath1 is not a good quantum number , though one can still imagine that it is a reasonable mean to count the modes of low enough frequency .
this pseudo - dispersion curve must have a slope for @xmath490 that corresponds to the transverse velocity .
in fact , as shown in fig .
4a , this slope well corresponds to the @xmath490 limit of the transverse velocity ( dashed line ) derived from fig . 2 .
on increasing @xmath1 , the data in fig .
4a show an early departure from linearity and a clear inflection at @xmath50 2 which is the frequency where the boson peak appears ( fig .
this inflection has been interpreted in terms of the failure of the classical born approximation for the description of the continuum elasticity in disordered systems @xcite .
it appears at the same frequency where the molecular dynamics results for the phase velocities in fig . 2 and fig .
3 show a softening .
this is not surprising as the acoustic - like frequencies can indeed be regarded as average values over a distribution of eigenvalues related to the broadening of the brillouin peaks . in other words ,
the velocity decrease observed at frequencies around the boson peak position in the acoustic - like excitations of transverse polarization ( fig .
2a ) and , through the shear component of the longitudinal modulus , in those of longitudinal polarization ( fig .
3a ) , directly reflects an inflection that appears in the @xmath1-dependence of the low-@xmath1 transverse eigenvalues .
as far as we can consider @xmath1 a reasonable quantum number , it is easy to directly estimate the vibrational density of states starting from the knowledge of the dispersion curves . in a simple plane - wave approach : @xmath51 \nonumber\\ & \simeq & \frac{3}{q_d^3 } \bigl(\frac{c_t}{c_d}\bigr)^3 \bigl(\frac{q^2}{\omega^2 } \frac{\partial q}{\partial \omega } \bigr)_t , \label{eq3}\end{aligned}\ ] ] where @xmath52 and @xmath20 stand for the longitudinal and transverse polarization , and
@xmath53 is the debye velocity .
the approximation in eq .
3 is justified by the fact that the longitudinal contribution to the total vibrational density of states is small since it scales with @xmath54=4% .
it is clear that in order to follow this approach we shall use the pseudo - dispersion curve obtained from the normal modes analysis since in that case the low - frequency eigenvalues are reasonably close to being plane waves . in order to compute eq .
[ eq3 ] , instead of directly differentiating the dispersion curve , we choose to use an empirical model to fit the simulation data ( full line through the points in fig .
4a ) and then to differentiate the model function .
the result of this calculation is shown in fig .
4b ( full line ) and , as it can be appreciated , it well describes the boson peak .
it is worth underlining that no additional input but the knowledge of the dispersion curve is required to perform this calculation .
it is then clear that the reduced density of states of the studied glass , up to the boson peak , originates from a deformation of the pseudo - dispersion curve of eigenstates that can still be approximately described using plane waves .
is reported as a function of frequency for both the transverse ( circles ) and longitudinal ( squares ) polarization .
the frequency where this ratio equals one defines the ioffe - regel limit . for the transverse case
, this limit falls very close to the boson peak position , indicated by the vertical dashed line .
different is the case of the longitudinal polarization , where the ioffe - regel limit appears at a frequency @xmath0 3 times higher than the boson peak position .
, scaledwidth=48.0% ] the molecular dynamics results presented above also allow us to calculate the ioffe - regel limit for the acoustic - like excitations , defined as the frequency where @xmath55 for both polarizations @xcite .
this limit corresponds to the frequency where the decay time of the acoustic - like excitations first matches half of the corresponding oscillation period , and therefore marks somehow an extreme upper bound in frequency for the validity of a plane - waves approach as a starting point to describe the acoustic - like excitations . as shown in fig . 5
, the present data confirm that the ioffe - regel limit for the transverse excitations is located close to the boson peak position @xcite .
5 also shows that the ioffe - regel limit is reached at different frequencies for the longitudinal and transverse excitations , and that the longitudinal one shows up at a frequency higher than the boson peak position @xcite .
however , it is important to underline that this last result seems not to be general : in a simulation study of a silica glass the ioffe - regel crossover was found to appear at the same frequency for both polarizations @xcite .
the connection between the ioffe - regel limit for the transverse excitations and the boson peak @xcite is clarified by the discussion above , which shows that this limit is reached in fact in the frequency range where the continuum approximation for the acoustic excitations breaks down .
summing up , the present simulation results shed new light on the well known universal anomaly observed in the specific heat of glasses in the @xmath56 10 k temperature range and related to the boson peak in the vibrational density of states at frequencies of @xmath01 thz .
we have shown that in glasses the elastic continuum approximation for the acoustic - like excitations breaks down abruptly on the mesoscopic , medium - range - order length - scale of about ten interatomic spacings , where it still works well for the corresponding crystalline systems .
this breakdown is signaled by a deformation of the pseudo - dispersion curve and corresponds to a marked reduction of the sound velocity on the mesoscopic scale .
this turns out to be the closely related to the aforementioned anomalies in the specific heat and vibrational density of states , that can be finally traced back onto elastic properties specific to glasses .
finally , in order to put the present results in some perspective and in connection with experiments , it is important to emphasize once more that they refer to a simple monatomic lj glass quenched with a cooling rate out of reach experimentally in other words there is no experimental analogue of the glass studied here .
still , we believe that such a simple model system has the great advantage of clearly grasping fundamental features that though observed in bits and pieces in many experiments are often hidden by a number of additional effects in real glasses . this
we believe is the reason why the vibrational properties of glasses are still a debated issue after several decades of studies and discussions .
for example , the high - frequency vibrational dynamics of the lj monatomic glass has been proven to be well described within the harmonic approximation @xcite . however ,
clearly real glasses are anharmonic systems in what cases will anharmonicity start to play a role there ?
moreover , differently from the investigated model , real glasses are often characterized by the presence of intramolecular or optic - like modes this issue is of course ( and on purpose ) completely disregarded here .
all of these questions will require studies on further and more complex models in order to be fully addressed .
we thank a. tanguy for discussions .
the present numerical investigation has been performed using simulation boxes of different sizes containing up to @xmath5710@xmath12 atoms , interacting via a lennard - jones potential with a cutoff @xmath582.5 and periodic boundary conditions .
this has been realized using the large scale massively parallel molecular dynamics computer simulation code lammps @xcite . a standard microcanonical classical molecular dynamics simulation , carried out at the constant number density @xmath59 1.015 and at temperature @xmath60 2 in the normal liquid phase is followed by a fast quench ( @xmath614@xmath6210@xmath28 ) down to @xmath20= 10@xmath21 .
the quenched glass sample is relaxed for a time ( dependent on the sample size ) sufficient to have a constant total energy .
the atomic positions , @xmath63 , and velocities , @xmath64 , have then been stored for a time ( again dependent on the sample size ) sufficient to get the desired resolution function .
the time correlation functions required to obtain the dynamic structure factor , @xmath14 , and its analogous function for the transverse excitations , @xmath13 have been computed as : @xmath65 where @xmath66 is @xmath52 or @xmath20 and : @xmath67\widehat{\mathbf{q } } \ ; e^{i \mathbf{q}\cdot\mathbf{r}_i(t ) } , \\
\mathbf{j}_{t}(q , t)&= & \sum_{i=1}^{n}\left\ { \left[\mathbf{v}_i(t)\cdot \widehat{\mathbf{q}}\right ] \widehat{\mathbf{q } } - \mathbf{v}_i(t ) \right\ } \;e^{i \mathbf{q}\cdot\mathbf{r}_i(t ) } , \label{eq5}\end{aligned}\ ] ] with @xmath68 . addressing the issue of the dependence of the obtained results on the quenching rate is a difficult task for the studied system due to the fact that it easily crystallizes .
the smallest quenching rate compatible with the long simulation times needed to reach the desired resolution and with the large sizes of the simulation boxes is about 5 decades smaller ( @xmath61 4@xmath6210@xmath21 ) than the one used for this study .
we have checked that the molecular dynamics results reported here are independent of the quenching rate in this range . for the glass studied here a standard normal modes analysis
has been carried out to derive the vibrational density of states , @xmath6 , from the eigenvalues of the dynamical matrix calculated in the inherent structures of the glass . moreover , the pseudo - dispersion curve for the transverse eigenvalues of the system reported in fig .
4a has been constructed following ref .
more in detail , for each of the considered simulation boxes , we collected the four lowest frequency degenerate eigenvalues obtained from the diagonalization of the dynamical matrix .
these modes correspond to the largest wavelength standing waves for the simulation box .
since the transverse sound velocity is @xmath692.3 times smaller than the longitudinal one , these four degenerate eigenvalues are all of transverse polarization .
for example , the first eigenvalue has a degeneracy of 12 and can be associated to wavevectors of the ( @xmath701,0,0 ) family ; the second eigenvalue has a degeneracy of 24 and can be associated to the wavevectors of the ( @xmath701,@xmath701,0 ) family , and so on .
these eigenvalues are size dependent : performing this analysis on simulation boxes of larger and larger size ( up to n=256000 particles ) , we then selected transverse eigenvalues with smaller and smaller @xmath1 s . a quite broad range of frequencies and @xmath1 s could thus be explored .
this procedure can be expected to lead to reasonable results only as far as the degeneracy of the eigenvalues shows up clearly enough to suggest a one - to - one relation to the corresponding values for @xmath1 , and becomes less and less reliable on decreasing the box size or on increasing @xmath1 . in the present case ,
we find that the highest @xmath1 values up to which this analysis is still reasonable are @xmath71 0.2 - 0.3 , corresponding to @xmath72 2 , i.e. basically up to the boson peak position , and becomes less and less reliable on further increasing @xmath1 . | the low - temperature thermal properties of dielectric crystals are governed by acoustic excitations with large wavelengths that are well described by plane waves .
this is the debye model , which rests on the assumption that the medium is an elastic continuum , holds true for acoustic wavelengths large on the microscopic scale fixed by the interatomic spacing , and gradually breaks down on approaching it .
glasses are characterized as well by universal low - temperature thermal properties , that are however anomalous with respect to those of the corresponding crystalline phases .
related universal anomalies also appear in the low - frequency vibrational density of states and , despite of a longstanding debate , still remain poorly understood . using molecular dynamics simulations of a model monatomic glass of extremely large size , we show that in glasses the structural disorder undermines the debye model in a subtle way : the elastic continuum approximation for the acoustic excitations breaks down abruptly on the mesoscopic , medium - range - order length - scale of about ten interatomic spacings , where it still works well for the corresponding crystalline systems . on this scale ,
the sound velocity shows a marked reduction with respect to the macroscopic value .
this turns out to be closely related to the universal excess over the debye model prediction found in glasses at frequencies of @xmath01 thz in the vibrational density of states or at temperatures of @xmath010 k in the specific heat . |
[ cols="^,^,^,^,^,^,>,<,^,^ " , ] the nebular emission line ratios can also probe the ionisation by an agn , by comparing the line ratios with , for example , a baldwin - philips - terlevich ( hereafter bpt ) diagram ( baldwin , phillips & terlevich 1981 ) .
this diagram separates star - forming galaxies and agn according to the optical line ratios @xmath2 6583/ and @xmath2 5008/. the star - forming galaxies fall in a region of lower @xmath2 6583/ for a range of values of @xmath2 5008/. the bordering line between the star - forming and agn regions are defined from kauffmann et al .
( 2003 ) that demarcates star - forming galaxies and agn on an empirical basis , and from kewley et al .
( 2001a ) which represents a limit to the line - flux ratios that can be produced for star - forming regions from photoionisation plus stellar population synthesis models .
recently , shapley et al .
( 2005 ) and erb et al .
( 2006a ) have presented evidence indicating a difference between regions in high - redshift galaxies and those in local galaxies .
they demonstrate that a fraction of the high - redshift star - forming galaxies at z@xmath312 lies offset from the local population of regions and star - forming galaxies , displaced toward higher @xmath2 6583/ and @xmath2 5008/ values . out of the three objects presented here , one has a @xmath2 6583 detection and the other two lead to lower limits .
we derive @xmath4 ( @xmath2 6583/ ) @xmath50.8 , = @[email protected] and @xmath50.6 for the absorbing galaxies towards q0302@xmath6223 , q0452@xmath61640 and q2352@xmath60028 , respectively .
similarly , both the @xmath2 5008 and lines are not detected in absorbing - galaxy towards q2352@xmath60028 .
we derive @xmath4 [ ( @xmath2 4960 + @xmath2 5008)/ ] = [email protected] and [email protected] for the absorbing galaxies towards q0302@xmath6223 and q0452@xmath61640 , respectively .
we find that all three objects do not show contribution from agns and prove to be consistent with starbursts .
even for q2352@xmath60028 which only has a limit from the @xmath2 6583/ ratio , the result indicates that it is consistent with expectations for star - forming galaxies with no apparent agn contamination or shock ionisation . in the local universe ,
disc galaxies almost invariably show negative metallicity gradients , with the central abundances being higher than the one in the outskirts of the galaxy .
the metallicity gradients in local galaxies are strongly correlated with hubble type , bar strength , and merging events , with more flattened gradients for barred galaxies and merging pairs ( kewley et al .
2010 ; rupke , kewley & chien 2010 ) .
the small sizes of high - redshift galaxies ( law et al .
2007 ; conselice et al .
2008 ; smail et al .
2007 ; swinbank et al .
2010 ) mean that sub - kpc resolution observations would be preferred to measure the abundances reliably . to overcome this problem , spatially resolved measurements of
/ have been made in gravitationally amplified galaxies at z=1.52.5 .
jones et al . ( 2010 ) measure strong , negative gradients at z@xmath32 suggesting that if z @xmath32 starbursts evolve into spiral discs today , then the gradients must flatten by 0.050.1 dex / kpc over the last 810gyr . furthermore , the high - redshift spiral galaxy metallicity gradient of yuan et al .
( 2011 ) is consistent with inside - out disc formation model .
more recently , swinbank et al . (
2012 ) have reported metallicity gradients in unlensed systems at 0.84@[email protected] which are negative or consistent with zero , with an average @[email protected] dex / kpc , which is comparable with the slope seen in the thick disc of the milky - way .
the authors find that all three isolated galaxies have negative gradients , while the merging system has a positive gradient .
in fact , only a few positive gradients have been measured , all of which are relatively shallow ( @xmath9 0.05 dex / kpc ) .
positive metallicity gradients have been claimed for some high redshift galaxies , most of them with no signs of recent interaction or merging .
cresci et al .
( 2010 ) and queyrel et al . (
2012 ) have suggested they could be caused by high accretion rates of metal - poor gas onto the nucleus which occur only at high - redshift or during mergers .
queyrel et al . (
2012 ) argue that these inflows are preferably due to interactions whereas cresci et al .
( 2010 ) relate them to cold flows .
queyrel et al . (
2012 ) also claim evidence for gas inflow and metallicity gradient flattening in close pairs and interacting galaxies .
these authors point to a transition period at z@xmath312 between cold accretion ( at high - redshift ) and mergers ( at lower - redshift ) in the growing process of massive galaxies .
these reported inverted gradients are puzzling since we know it is difficult to form a positive gradient given the fraction of gas converted into stars per unit time must increase with radius , while the dynamical time - scale generally decreases with radius .
the corollary is that a higher fraction of metal - enriched gas must be lost to outflows at small radii .
positive gradients for interacting systems may be a signature of the redistribution of the metal - rich gas produced from a central starburst ( werk et al .
2010 ; rupke , kewley & chien 2010 ) , or possibly as a result of inflow of relatively unenriched gas from the halo ( or igm ) if it is able to intersect the central regions of the disc galaxy without being disrupted .
another explanation is an increasing contribution of shocks at large radius rather than a positive metallicity gradient ( kewley et al . 2010 ) .
we now know how merging processes affect the metallicity gradients : the gradients flatten immediately after first close pass ( barton , geller & kenyon 2003 ; kewley , geller & barton 2006 ) .
the metallicity in mergers can be used as a signpost of first pericentre passage ( kewley et al .
models of kobayashi ( 2004 ) predicts that in the case of monolithic collapse , there will be steep metallicity gradient which shows no time evolution . on the contrary ,
major mergers are expected to flatten the metallicity gradient with signs of an evolution with time .
therefore , the gradient evolution with redshift is related to the radial size growth predicted by models . recently , swinbank et al .
( 2012 ) reproduced their observed metallicity gradients with predictions for the gas discs of star - forming galaxies in the cosmologically based hydrodynamic simulations . in these models
, the inner disc undergoes initial rapid collapse and star - formation with gas accretion ( from the halo and/or igm ) depositing relatively unenriched material at the outer disc , causing negative abundance gradients . at lower redshift ( when gas accretion from the igm is less efficient )
the abundance gradients flatten as the outer disc becomes enriched by star - formation ( and/or the redistribution of gas from the inner disc ) .
we note however that m101 is an important exception to this description , being at low - redshift but with a strong gradient ( kennicutt , bresolin & garnett 2003 ) .
the systems discussed in this paper are part of a larger group of galaxies observed with sinfoni . for three of
the absorbing - galaxies observed in that sample , we have built 2d metallicity maps based on detection .
the map of the system towards q1009@xmath60026 is presented in proux et al .
( 2011a ) and the ones for q0452@xmath61640 and q2352@xmath60028 studied here are presented in proux et al .
( 2013a ) .
we derive internal metallicity gradients of @[email protected] , @[email protected] and @[email protected] dex / kpc for q0452@xmath61640 , q1009@xmath60026 and q2352@xmath60028 , respectively .
therefore , the internal gradients of metallicity with radius of these three objects are flat or at most weakly negative for two of the three systems .
these estimates are in line with the gradients reported in the local universe .
for the local galaxy gradients , the isolated spiral control sample of rupke , kewley & chien ( 2010 ) indicates a median slope of @[email protected] dex / kpc .
the median slope of gradients from that sample is consistent with the values of zaritsky , kennicutt & huchra ( 1994 ) and van zee et al .
( 1998 ) samples .
the zaritsky , kennicutt & huchra ( 1994 ) gradient sample is composed of 39 local disc galaxies , and the van zee et al .
( 1998 ) sample added another 11 local spirals .
local early - type galaxies have even shallower gradients ( henry & worthey 1999 ) .
therefore , our observations resemble more the local samples or the results of queyrel et al .
( 2009 ) at [email protected] than the gradient recently reported in lensed galaxies at z@xmath32 ( jones et al . 2010 ; yuan et al . 2011 ; jones et al .
2012 ) .
prantzos & boissier ( 2000 ) modeled metallicity gradients of spiral galaxies for different velocities ( v@xmath10 ) and spin parameters ( @xmath2 ) .
we find that the metallicity gradients reported here are comparable with the predictions by prantzos & boissier ( 2000 ) for v@xmath10 = 220 km / s which is typical of the dynamics of these systems ( see proux et al .
2013a ) . at z @xmath31 , the model disc galaxies from the gimic simulation presented by sinwbank et al .
( 2012 ) display negative metallicity gradients which are also consistent with our observations .
abundances of galaxies are typically inferred from oxygen in nebular ( i.e. emission line ) spectra of regions .
it is important to note that region may suffer from self - enrichment ( kunth & sargent 1986 ) and the bulk of metals in star forming galaxies may in fact be in the neutral gas ( up to 9095% ) .
it is therefore important to be able to compare these two metallicity indicators both in the local universe and at high - redshift . in the milky way , the oxygen abundances in orion s regions ( bautista & pradhan 1995 ) are in good agreement with those measured from uv absorption lines in the local interstellar medium ( moos et al .
2002 ) . in the past years
, izw18 has been studied in great details ( aloisi et al .
2003 ; lecavelier des etangs et al .
2004 ; lebouteiller et al .
2004 ) using fuse observations .
the neutral gas is found to be already enriched in heavy elements , i.e. not primordial .
the @xmath0-elements traced by o , ar and si are @xmath3 0.5 dex lower in the neutral interstellar medium ( ism ) but the fe abundance is found to be similar to the ionised phase , probably because fe is release on a longer time - scale . a star formation at least @xmath3 1 gyr old is therefore required to explain the fe abundance of the ism .
lebouteiller et al .
( 2009 ) studied a sample of low metallicity galaxies and find that abundances in neutral gas are equal or lower than in the ionised gas and confirm a plateau below which metals are not found .
we note that these results seem to put into question the reliability of regions to be a good tracer of the ism metallicity , but the difference in abundance between the neutral and ionised gas may be a by - product of the sample selection . other studies at z=0 are underway ( james et al .
2011 ) for 9 galaxies based on acs and cos observations .
the authors find that overall the abundances are similar to those in region except for n , higher in regions , ( suggesting enrichment in regions by wolf - rayet stars ) .
the oxygen metallicity in and agree well .
early results indicate that the difference between distinct gas phase are smaller towards the nucleus of the galaxies probably because there is more efficient mixing .
these findings imply that we might indeed expect emission and absorption abundances to be in good agreement , at least on scales of a few kpc . similarly , schulte - ladbeck et al .
( 2004 , 2005 ) compare stis abundances of the dla towards the quasar hs 1543@xmath115921 and the emission line abundances of the region in the dwarf spiral sbs 1543@xmath11593 . in that case , the angular separation is only @xmath33kpc . the authors find that the neutral and ionised gas abundances agree , maybe because of more efficient mixing from rotation in that spiral .
bowen et al . ( 2005 ) utilised all the hst / stis data obtained for this object and were able to confirm these conclusions . for the quasar absorption line galaxies in our sample , both and metallicities free from dust - bias
are available .
this allows us to compare the metallicity at the centre of the galaxy ( seen in emission ) with the metallicity at a radius of a few dozen of kpc corresponding to the impact parameter ( seen in absorption ) . at present
, there are only nine systems ( including five from our recent publications ) for which both these measurements are available . for the three of these (
q0302@xmath6223 , q0452@xmath61640 and q2352@xmath60028 ) , we have presented here robust metallicity estimates based on several emission line indicators observed with x - shooter and reported in table [ t : hii_metals ] .
in all three cases , the metallicity derived for the gas is slightly lower than the one measured in the ionised gas .
if we use these data to derive abundance gradients over 25 kpc , 16 kpc and 12 kpc ( see table 1 ) , we obtain @[email protected] , @[email protected] and @xmath50.04 dex / kpc for q0302@xmath6223 , q0452@xmath61640 and q2352@xmath60028 , respectively . these slopes depict flat gradients . in the three cases where
is detected in the sinfoni data , we can compare the internal gradients derived from 2d information within the galaxies in the section above with the gradients deduced from a comparison of the ionised and neutral phases .
we recall that we derived @[email protected] , @[email protected] and @[email protected] dex / kpc for the ionised gas abundance within the galaxies , while proux et al .
( 2012 ) reported @[email protected] , @[email protected] , and @xmath50.04 dex / kpc over 16 kpc , 39 kpc and 12 kpc between the and phases for q0452@xmath61640 , q1009@xmath60026 and q2352@xmath60028 , respectively .
while these gradients are spanning different physical scales , they agree within the admittedly large error bars .
this suggests that neutral gas abundances are good tracers of the metallicity of a galaxy .
since the internal gradient within the ionised gas aligns well with the gradient between the ionised gas in the nucleus to the neutral gas in the halo , one can assume that the latter is an acceptable proxy of the former .
however , this assumes that the chemical evolution of the neutral gas mirrors that of the ionised gas , i.e. that a gradient in the neutral gas also exists .
by analogy with studies of local galaxies , we expect interstellar dust to be a pervasive component of gas of galaxies observed in absorption along the line - of - sight to the background quasar .
the amount of dust is empirically inferred from the relative abundances of iron - peak elements with respect to zinc , which is not depleted much on to dust grains in the interstellar medium of the milky way ( meyer & roth 1990 ) . for the dla towards q0302@xmath6223 , the elemental ratios [ fe / zn][email protected]@xmath70.17 ,
[ cr / zn][email protected]@xmath70.17 and [ mn / zn][email protected]@xmath70.17 indicate a rather large amount of depletion in the system with respect to the mean values for dlas ( noterdaeme et al .
pettini et al .
( 2000 ) note a strong difference in the dust content of the two main components in this system : the component at v=0 km / s has [ fe / zn][email protected] , [ cr / zn][email protected] and [ mn / zn][email protected] while the one at v=@xmath636 km / s has [ fe / zn][email protected] , [ cr / zn][email protected] and [ mn / zn][email protected] .
the authors recall that in the local interstellar medium the so - called routly - spitzer effect ( routly & spitzer 1952 ) also indicates reduced depletions in high - velocity interstellar clouds . for the dla towards q0452@xmath61640 , the elemental ratios [ fe / zn][email protected]@xmath70.11 , [ cr / zn][email protected]@xmath70.12 and [ mn / zn][email protected]@xmath70.14 indicate a rather modest amount of depletion . for the sub - dla with log = 19.81@xmath12 towards q2352@xmath6002 , the elemental ratios
[ fe / zn]@xmath130.14 and [ cr / zn]@xmath130.02 indicate a rather small amount of depletion , atypical of the lower systems ( meiring et al .
2009 ) .
the x - shooter spectra presented here allow for a detailed study of the dust content of the regions in the absorbing - galaxies .
extinction in z@xmath142 star - forming galaxies is commonly estimated from rest - frame uv colors and an application of the calzetti et al .
( 2000 ) starburst attenuation law .
the calzetti law appears to provide a fairly accurate description on average of the reddening and attenuation of the uv stellar continuum in both nearby and distant starburst galaxies ( reddy & steidel 2004 ; reddy et al .
the degree of dust extinction in star - forming regions can also be estimated from the balmer lines of hydrogen , because these strong optical lines have intrinsic ratios that are well described by atomic theory . under the assumption of case b recombination ( osterbrock 1989 ) and for t
= 10 , 000 k , the balmer ratios are set .
any deviations in the observed line ratios are then attributed to dust extinction .
calzetti ( 2001 ) demonstrates that , in local star - forming galaxies , the stellar continuum suffers less reddening than the ionised gas , expressed as @xmath15 .
furthermore , calzetti ( 2001 ) suggests that the reddening of the balmer lines in nearby uv - selected starbursts is better described by a foreground dust distribution and traditional milky way extinction curve ( cardelli , clayton & mathis 1989 ) . on the other hand , erb et al .
( 2006b ) present evidence that , in z@xmath32 star - forming galaxies , @xmath16 and a calzetti et al .
starburst extinction law applied to both uv - continuum and emission lines gives rise to the best agreement between uv- and -derived sfrs . in the case of quasar absorbers ,
york et al .
( 2006 ) have used a sample of more than 800 absorption systems with 1@xmath8z@xmath17@xmath82 in the sloan digital sky survey ( sdss ) to show that the average extinction curve in these systems is similar to the smc extinction curve , with no 2175 feature and a rising uv extinction curve below 2200 .
we note , however , that a small number of absorbers have been reported to have the 2175 features ( e.g. junkkarinen et al .
2004 ; wang et al . 2004
; srianand et al .
2008 and kulkarni et al .
2011 ) . for larger systems traced by dlas ,
the evidence for extinction are weak ( frank & proux 2010 ; khare et al .
2012 ) , which is possibly a limitation intrinsic to the sdss survey and quasar - selection itself ( noterdaeme et al .
2009 ) . here
, we use the detection of both and in the absorbing - galaxy to compute the balmer decrement , which can allow us to infer the dust extinction by comparing balmer line ratios to their predicted values via atomic physics .
the colour excess can be computed from the following equation : @xmath18 these values are the intrinsic / , and k(@xmath2 ) is derived assuming an extinction curve of the smc at wavelength @xmath2 .
we derive the parameters k ( @xmath19 ) and k ( @xmath20 ) from an interpolation of the empirical extinction curves of pei ( 1992 ) and the relation k(@xmath2 ) = a(@xmath2)/@xmath21 from calzetti ( 2004 ) as described in gharanfoli et al .
we estimate the extinction correction assuming a milky way type ( mw - type ) law .
the resulting gas - phase colour excesses are estimated to be @[email protected] , [email protected] and @xmath220.4 for the absorbing - galaxies towards q0302@xmath6223 , q0452@xmath61640 and q2352@xmath60028 , respectively .
these values are listed in the last column of table [ t : hii_metals ] .
these estimates have an opposite trend to the values derived from abundance ratios in the neutral gas phase in the section above .
indeed , the system towards q0302@xmath62238 presents a higher level of extinction in absorption but a lower value of @xmath21 , while the absorbing - galaxy towards q0452@xmath61640 indicates a more modest amount of extinction in absorption but a higher value of @xmath21 .
thus , the gas directly in front of the quasar appears to have different dust properties than the absorbing - galaxy s gas probed at the line of sight in our emission - line spectroscopy . in the case of the object towards q2352@xmath60028 ,
the @xmath21 derived is minimally constrained due to the absence of detection of the emission line .
figure [ f : whole_spec ] presents the full x - shooter quasar spectra flux - calibrated in erg / s/@xmath23/ .
overplotted on each panel is the sdss quasar composite spectrum of vanden berk et al .
( 2001 ) redshifted to the object position and arbitrarily scaled to the observed fluxes .
because of the large wavelength coverage of the x - shooter observations , the spectra offer a strong constraint on the shape of the quasar continuum .
indeed , it is a common practice to compare the shape of an individual quasar spectrum with a composite to derive the reddening of the quasar ( fynbo et al .
2010 ; noterdaeme et al . 2012 ; fynbo et al .
this information is difficult to relate to the integrated effect of dust due the intervening absorbers , because it can not be differentiated from the intrinsic colour of the quasar which is known to vary from one object to another .
a more robust approach to an estimate of the reddening from the measure of the quasar continuum slopes has to come from a statistical approach to overcome the inherent difference from one object to another ( murphy & liske 2004 ; york et al .
2006 ; frank & proux 2010 ; khare et al . 2012 ; kulkarni et al .
we have presented x - shooter emission - line spectra covering the observed wavelength 300 nm to 2.5 @xmath1 m for three absorbing - galaxies reported by proux et al .
( 2011a , 2012 ) .
the slit was oriented to obtain a spectrum of both the background quasar and the galaxy responsible for the high- quasar absorber .
these data thus allow for a robust estimate of abundance in these and allow for a comparison of the metallicities in the neutral and ionised phase of the same objects .
our results suggest that the abundances derived in absorption along the line - of - sight to background quasars are a reliable measures of the overall galaxy metallicities .
the 2d metallicity maps based on sinfoni observations previously published ( proux et al .
2011a , 2013a ) , show small negative metallicity gradients .
the flat slopes are in line with the differences observed between the two phases of the gas .
these results suggest that a comparison of the and metallicities is a robust indicator of the internal gradients .
finally , we use several indicators to measure the quantity of dust in the ionised and neutral phases of these systems . the presence of dust in the phase
can be estimated from the observed depletion of refractory elements .
we find measures from the balmer decrement of the galaxies to be in line with values derived from element ratios in the gas .
we note however , that the smaller column density towards q2352@xmath60028 show little indication of dust depletion , contrary to what is expected from such sub - dlas .
we are grateful to nicolas bouch for his contribution to the overall project .
we would like to thank the paranal and garching staff at eso for performing the observations in service mode and the instrument team for making a reliable instrument .
vpk acknowledges partial support from the u.s .
national science foundation grants ast/0908890 , ast/1108830 ( pi : kulkarni ) .
this work has benefited from support of the agence nationale de la recherche with reference anr-08-blan-0316 - 01 .
epinat , b. , tasca , l. , amram , p. , contini , t. , le fvre , o. , queyrel , j. , vergani , d. , garilli , b. , kissler - patig , m. , moultaka , j. , paioro , l. , tresse , l. , bournaud , f. , lpez - sanjuan , c. & perret , v. , 2012 , a&a , 539 , 92 fynbo , j. p. u. , laursen , p. , ledoux , c. , moller , p. , goldoni , p. , gullberg , b. , kaper , l. , maund , j. , noterdaeme , p. , ostlin , g. , strandet , m. l. , toft , s. , vreeswijk , p. m. & zafar , t. , 2010 , mnras , 408 , 2128 fynbo , j. p. u. , ledoux , c. , noterdaeme , p. , christensen , l. , moller , p. , durgapal , a. k. , goldoni , p. , kaper , l. , krogager , j. -k . ,
laursen , p. , maund , j. r. , milvang - jensen , b. , okoshi , k. , rasmussen , p. k. , thorsen , t. j. , toft , s. & zafar , t. , 2011 , mnras , 413 , 2481 | the gas - phase and stellar metallicities have proven to be important parameters to constrain the star formation history of galaxies . however , regions associated with recent star - formation may not have abundances typical for the galaxy as a whole and it is believed that the bulk of the metals may be contained in the neutral gas .
it is therefore important to directly probe the metal abundances in the neutral gas , which can be done by using absorption lines imprinted on a background quasar .
recently , we have presented studies of the stellar content of a small sample of such quasar absorbers with column densities measured to be in the sub - damped lyman-@xmath0 to damped lyman-@xmath0 range . here
, we present observations covering 300 nm to 2.5 @xmath1 m of emission line spectra of three of these absorbing - galaxies using the long - slit spectrograph x - shooter on the vlt .
this allows us to compare the neutral and ionised phase metallicities in the same objects and relates these measures to possible signature of low - metallicity gas accretion or outflows of gas enriched by star formation .
our results suggest that the abundances derived in absorption along the line - of - sight to background quasars are reliable measures of the overall galaxy metallicities .
in addition to a comparison of abundances in different phases of the gas , a potential observational consequence of differences in fueling mechanisms for disc galaxies is the internal distribution of their chemical abundances .
we present some evidence for small negative metallicity gradients in the three systems .
the flat slopes are in line with the differences observed between the two phases of the gas .
these results suggest that a comparison of the and metallicities is a robust indicator of abundance gradients in high - redshift galaxies and do not favour the presence of infall of fresh gas in these objects .
@xmath2 [ firstpage ] galaxies : formation galaxies : evolution galaxies : abundances galaxies : ism quasars : absorption lines intergalactic medium |
detectors and instrumentation have always been the biggest drivers of sensitivity improvements in astrophysics .
while the collecting area of ground - based telescopes has only improved by a factor of 4 in the last 50 years , the per - pixel sensitivity of astronomical detectors has grown by at least a factor of 20 , and the improvements in quality and size have been even more impressive .
these improvements have led to a golden age in astrophysics , but we are rapidly reaching a plateau in the per - pixel performance of traditional semiconductor based ( ccd @xcite and hgcdte @xcite ) detectors , with improvements only coming in the total size and pixel count of the final mosaic @xcite . to maintain the rapid pace of advance in per - pixel performance of detectors for astrophysics
requires breaking away from semiconductors - based detectors , and developing technologies with potentially greater performance ; for instance by reducing read noise and dark current , having a wider bandwidth , or having inherent spectral resolution .
the most promising avenue is to use superconducting detectors @xcite these detectors significantly increase performance by operating below 4 kelvin , thereby reducing the contribution from thermal noise , allowing them to measure the energy and arrival time of a single photon with no false counts .
the * * ar**ray * * c**amera for * * o**ptical to * * n**ear - ir * * s**pectrophotometry ( arcons ) was built around a highly multiplexible type of low temperature detector ( ltd ) known as microwave kinetic inductance detectors ( mkids ) @xcite , optimized for optical and near - infrared astronomy .
the goal of arcons is to demonstrate the viability of both the mkid technology as well as do science that would be difficult or impossible with conventional instruments .
an mkid array gives arcons significant advantages over a conventional lenslet , fiber fed , or image slicer integral field spectrograph ( ifs ) such as : * simple optical design that enables very high throughput * time resolution up to six orders of magnitude better than a ccd * extremely broad intrinsic bandwidth ( 0.15 @xmath1 m ) with good quantum efficiency * no read noise or dark current , and nearly perfect cosmic ray rejection * no observing time is lost to readout of the array * simpler scaling to much larger arrays than conventional ifss .
for example , harmoni @xcite for the e - elt will use 8 4k@xmath04k detectors but only has 32768 spaxels . in an mkid ,
each pixel is a spaxel , albeit at significantly lower spectral resolution than an ifs like harmoni .
* time domain information allows _ ex post facto _ use of calibration stars for monitoring atmospheric transparency , setting dynamic apertures , and applying tip / tilt corrections .
this improves signal to noise and calibration accuracy , and allows the instrument to make optimal use of the atmospheric conditions .
* photon arrival time , spectral resolution , and the large number of pixels allow for monitoring and the removal of sky emission , potentially down to the poisson limit , even in spectral regimes dominated by time variable oh emission the mkid technology in arcons is relatively immature , leaving significant room to improve future instruments .
current optical mkid technology has a spectral resolution r=@xmath2 at 4000 .
improvements in the mkid and readout electronics are on track to increase the spectral resolution towards the theoretical limit of around 100 for a 100 mk operating temperature . the array size , though small compared to an imaging ccd ( although not small compared to ifss ) , is currently only limited by what we can afford to readout with our room temperature electronics ( section [ sec : readout ] ) . as the components used in these electronics are improving with moore s law , megapixel arrays should be possible within a decade .
we describe the mkid arrays in detail in section [ sec : mkid ] .
the optical design is presented in section [ sec : optical ] , and the instrument design , including cryogenics , mounting hardware , guide camera , and wavelength calibration system is presented in section [ sec : inst ] .
section [ sec : readout ] describes the cryogenic and room temperature readout electronics , and section [ sec : pipeline ] contains a brief outline of the data storage and processing .
section [ sec : perf ] contains on - sky performance results .
low temperature detectors ( ltds ) with operating temperatures on the order of 100 mk , are currently the preferred technology for astronomical observations over most of the electromagnetic spectrum , notably in the far infrared through millimeter ( 0.13 mm ) @xcite , x - ray @xcite , and gamma - ray @xcite wavelength ranges . in the important ultraviolet , optical , and near - infrared ( uvoir ) ( 0.15
@xmath1 m ) wavelength range a variety of detector technologies based on semiconductors , backed by large investment from both consumer and military customers , has resulted in detectors for astronomy with large formats , high quantum efficiency , and low readout noise .
these detectors , however , are fundamentally limited by the large band gap of the semiconductor which restrict the maximum detectable wavelength ( 1.1 ev for silicon ) and thermal noise sources from their relatively high ( @xmath3100 k ) operating temperatures .
ltds allow the use of superconductors with gap parameters over 1000 times lower than semiconductors .
this difference allows a leap in capabilities .
a superconducting detector can count single photons with no false counts while determining the energy ( to several percent or better ) and arrival time ( to a microsecond ) of the photon ( the optical analog of an x - ray calorimeter ) .
it can also have much broader wavelength coverage since the photon energy is always much greater than the gap energy .
while a ccd is limited to about 0.31 @xmath1 m , the mkids described here are in principle sensitive from 0.1 @xmath1 m in the uv to greater than 5 @xmath1 m , enabling observations at important infrared wavelengths .
superconducting uvoir detectors have been pursued in the past with two technologies , superconducting tunnel junctions ( stjs ) @xcite and transition edge sensors ( tess ) @xcite . while both of these technologies result in functional detectors , they are limited to single pixels or small arrays due to the lack of a credible strategy for wiring and multiplexing large numbers of detectors , although recently there have been proposals for larger tes multiplexers @xcite .
microwave kinetic inductance detectors , or mkids @xcite , are a newer cryogenic detector technology that has proven important for astrophysics due to their sensitivity and the ease with which they can be multiplexed into large arrays .
the `` microwave '' in mkids comes from their use of frequency domain multiplexing @xcite at microwave frequencies ( 0.120 ghz ) that allows thousands of pixels to be read out over a single microwave cable .
the optical lumped element ( ole ) @xcite mkid arrays we have developed have significant advantages over semiconductor detectors .
they can count individual photons with no false counts and determine the energy and arrival time of every photon with good quantum efficiency . their physical pixel size and maximum count rate
is well matched with large telescopes .
these capabilities enable powerful new astrophysical instruments usable from the ground and space .
mkids work on the principle that incident photons change the surface impedance of a superconductor through the kinetic inductance effect @xcite .
the kinetic inductance effect occurs because energy can be stored in the supercurrent ( the flow of cooper pairs ) of a superconductor .
reversing the direction of the supercurrent requires extracting the kinetic energy stored in it , which yields an extra inductance term in addition to the familiar geometric inductance .
the magnitude of the change in surface impedance depends on the number of cooper pairs broken by incident photons , and is hence proportional to the amount of energy deposited in the superconductor .
this change can be accurately measured by placing a superconducting inductor in a lithographed resonator .
a microwave probe signal is tuned to the resonant frequency of the resonator , and any photons which are absorbed in the inductor will imprint their signature as changes in phase and amplitude of this probe signal .
since the quality factor @xmath4 of the resonators is high and their microwave transmission off resonance is nearly perfect , multiplexing can be accomplished by tuning each pixel to a different resonant frequency with lithography during device fabrication . a comb of probe signals can be sent into the device , and room temperature electronics can recover the changes in amplitude and phase without significant cross talk @xcite , as shown in the right panel of figure [ fig : detcartoon ] . the primary theoretical limitation on the spectral resolution is from the intrinsic quasiparticle creation statistics during the photon absorption event .
the energy from the photon can end up in two places , the quasiparticle system and the phonon system .
these systems interact , allowing energy exchange between the two , which reduces the statistical fluctuation from poisson by the fano factor @xmath5 , typically assumed to be 0.2 @xcite .
the spectral resolution of the detector , @xmath6 , can be written as @xmath7 , where @xmath8 is the efficiency of conversion of energy into quasiparticles , typically 0.57 @xcite , @xmath9 is the photon energy , @xmath5 is the fano factor , and @xmath10 is the energy gap .
the energy gap depends on the superconducting transition temperature @xmath11 of the inductor , @xmath12 , and we typically operate at a base temperature of @xmath13 . going to lower @xmath14 , and hence lower operating temperature ,
improves the theoretical @xmath15 .
operating at 100 mk yields a theoretical spectral resolution of r@xmath16100 at 400 nm .
previous research with superconducting tunnel junctions ( stjs ) with superconducting absorbers has shown that superconducting absorbers can approach the fano limit @xcite .
mkids are very versatile , as essentially any resonator with a superconductor as the inductor will function as a mkid . in 2008
we decided to pursue a lumped element resonator design @xcite .
the resonator itself consists of a 60 nm thick @xcite sub - stoichiometric titanium nitride ( tin@xmath17 ) film @xcite with the nitrogen content tuned with @xmath18 such that the superconducting transition temperature @xmath14 is about 800 mk . due to the long penetration depth of these films ( @xmath31000 nm )
the surface inductance is a high 25 ph / square , allowing a very compact resonator fitting in a 222@xmath19222 @xmath1 m square .
a square microlens array with a 92% optical fill factor is used to focus light onto the photo - sensitive inductor . the pixel pitch is easily controlled , with pitches between 75500 relatively easy to achieve .
the quasiparticle lifetime in our tin films is 50100 @xmath1s .
this sets the pulse decay time , allowing a maximum count rate of approximately 2500 cts / pixel / second before problems arise in separating pulses .
after significant development @xcite , we have arrived at the 2024 ( 44@xmath046 ) pixel array shown in figure [ fig : mkids ] .
typical performance for single pixel devices is r@xmath2010 at 4000 . over an entire array ,
we see a median spectral resolution r@xmath208 at 4000 ( section [ sec : res ] ) .
typically @xmath2190% of the resonators show up in frequency sweeps , but due to the variations in the tin gap there are a significant number of collisions ( two or more pixels with resonant frequencies closer together than 500 khz ) , reducing usable pixels , as discussed in section [ sec : yield ] . after cutting out collisions and pixels with especially high or low quality factor
, we usually have @xmath2070% of our pixels usable
. a more uniform film , such as the ones made at nist with ti / tin multilayers @xcite or with atomic layer deposition ( ald ) , could significantly reduce the number of collisions , dramatically improving yield .
even with the current moderate spectral resolution and yield , these arrays are some of the most powerful tools for narrow field of view astrophysics ever developed .
40 inductor .
various features of interest are labeled .
right : a zoomed in view of the array in the left panel . ]
an optical system , shown in figure [ fig : raytrace ] , has been designed with zemax to couple arcons with the f/30 beam at the palomar 200 " coud focus . a very similar system ( using just a different dewar entrance lens )
is used to couple to the lick 3-m coud focus .
this focus is stationary , which makes interfacing with the cryostat ( and its compressor ) significantly easier , and the time resolution of the mkids means that rotation of the image plane over time can be easily removed .
the optical path is as follows : light bounces off the primary and secondary , then off the coud m3 that directs light down the polar axis of the equatorial mount .
once in the coud antechamber , a 100 mm diameter pickoff mirror ( m4 ) sends the light towards the dewar entrance window .
before the dewar entrance , two pickoff mirrors form a @xmath2025 arcsecond wide slit and direct the rest of the field of view towards the guide camera s reimaging optics . immediately in front of the entrance window is a commercial thor labs 1 " filter wheel , which can hold up to 6 filters for calibration . observing with arcons
is typically done at a filter wheel position that does not contain a filter .
the dewar entrance window is an achromatic doublet which collimates the light .
all the powered optics have a edmund optics vis - nir anti - reflection coating ( reflection @xmath22% per surface between 4000 - 10000 ) except the bk7 meniscus lens ( surface 11 and 12 in table [ table : prescrip ] ) which is uncoated .
the light then passes through the dewar s 77 k shield though a baffle , and encounters the dewar s 4 k radiation shield .
at the 4 k radiation shield an asahi supercold filter is placed to set the red edge of arcons s wavelength sensitivity .
this was set conservatively at 1.1 in order to reduce sky count rates , but in subsequent runs we will attempt to extend the wavelength coverage out through j band to 1.35 . there is a pupil image at 4 k suitable for placing a cold 2.1 mm diameter lyot stop , but this turned out not to be necessary because the glass in the optics at 4 k blocks out enough of the 300 k thermal blackbody radiation to keep the 100 mk stage from warming up .
the supercold filter does effectively function as a cold lyot stop with a diameter of 12 mm .
there are two achromatic doublets at 4 k to decollimate the beam and set the plate scale , but these achromats are physically separated to eliminate any potential issues with differential thermal contraction between the different glasses . after the final 4 k optic , the light reaches the 100 mk stage . here the 4 k blackbody radiation is blocked out with a thin piece of borosilicate glass coated in indium tin oxide ( ito ) , a transparent conductor .
the glass absorbs infrared up to roughly 400 , and the ito reflects a substantial portion of the longer wavelength radiation .
finally , a square microlens with a 0.92 mm focal length and a 92% optical fill factor focuses the light on the sensitive mkid inductor .
the optical prescription is shown in tabular form in table [ table : prescrip ] .
the resulting spot diagram can be seen in figure [ fig : spot ] .
the simulation includes the microlens array in front of the detector .
the mkids pixel pitch of 222 means that the optical system does not need high magnification , simplifying the system .
the system is designed to have 0.45 " pixels on the sky .
the clear aperture of the optical system was designed to allow an unvignetted field of view of @xmath23 arcseconds .
the left panel shows the spot on axis , and the right panel shows the spot at the extreme right side of the array . even at the edges of the array
the vast majority of the spot is contained within the 40@xmath040 inductor . ]
this optical system was over - designed with every precaution possible to block out thermal blackbody radiation , which did not turn out to be required .
future mkid cameras will likely completely eliminate the powered cold optics in favor of much simpler optical systems , including just warm optics to adjust the plate scale , a flat dewar window , a 4 k bk7 flat window , and a 100 mk ito on bk7 ir blocker . [
cols="<,^,^,^,^,^,<",options="header " , ]
arcons is built around an adiabatic demagnetization refrigerator ( adr ) made by janis research .
the system contains a cryomech pt-405 two stage pulse tube cooler .
the first stage of the pulse tube cools the outer radiation shield to 5070 k , while the second stage cools the adr to 3.3 k. the pulse tube is powered by an air cooled compressor which can be located up to 100 from the cryostat . at palomar
, the compressor is placed in a small room on the telescope floor that has a hatch that can be opened to vent the heat outside of the dome . at lick
, the compressor is placed in the battery room outside the dome .
the adr is a single shot magnetic cooler . during the afternoon
the superconducting adr magnet is ramped to 40 kg , the heat switch is opened , and then the magnet is slowly ramped down during the course of the night .
a pid loop on a lakeshore 370 controller provides input to a lakeshore 625 superconducting magnet power supply , allowing us to stabilize the temperature of the adr to better than 50 @xmath1k .
stabilizing the temperature at 110 mk provides over 12 hours of hold time .
there are no obvious drifts in the mkids resonant frequencies as the magnetic field remaining in the superconducting magnet ramps down from 0.75 kg to 0 kg over the 12 hours hold time . at the palomar 200 "
, an aluminum frame is attached to a steel plate in the coud antechamber that is used to accommodate the coud spectrograph s slit jaws .
arcons hangs down from this aluminum frame , as shown in figure [ fig : arconsmount ] . at the lick 3-m
, arcons is also placed in the coud antechamber , but since the hamilton spectrograph is still in use the mounting plate is not used .
instead , arcons is hung from an i - beam in the ceiling and then stabilized to the mounting plate with a steel strut .
the guide camera for arcons is a sbig stf-8300 m ccd camera .
light is reflected off mirrors on either side of the arcons entrance aperture , bounces to an off - axis parabaloid , and then goes through a nikon 50 mm f/1.4 lens mounted to the sbig camera .
the unvignetted field of view of the guide camera is about 1.5 arcminutes in diameter . the sbig camera , using 3@xmath03 binning , is read out over usb by a small windows computer located in the electronics rack . in 10 second exposures we are able to guide off of stars in the 1820@xmath24 magnitude range . for brighter objects ,
exposure times down to 0.1 seconds are used .
this computer is accessed by remote desktop in the control room .
the guiding software is a modified version of the palomar observatory s standard guiding package . as part of the data taking procedure
each pixel s raw phase response must be mapped to photon energy .
we developed a system to uniformly illuminate the arcons focal plane with three lasers at the same time before and after science observations , thereby providing a reference to calibrate each detector s response .
the three lasers operate at 4066 , 6710 , and 9821 .
they are integrated into a package that allows for adjustment of the intensity of each laser s light .
the light from the three lasers is combined in an integrating sphere with a fiber optic output .
the fiber optic carries the combined light to optics and diffusers that simulate the f/30 telescope beam .
a mirror on a mechanical arm can be rotated into position to block out the sky light and direct the calibration beam into the dewar .
the amount of exposure time and frequency of calibrations were controlled through software . during observations ,
this process occurred approximately once every hour .
the response for a single detector is shown in figure [ fig:3peaks ] , where the number of counts for a one minute integration are plotted against detector phase in blue , using bins of width one digital number as returned by the firmware .
three gaussians were simultaneously fit to the three peaks produced in response to the laser light , with the width of each gaussian being a measure of the noise and therefore spectral resolution of the detector .
the fit is shown in red .
the tail out to low energy ( less negative phase response ) is due to the low energy background ( thermal blackbody radiation , noise triggers , etc . ) and is not included in the fit .
the location of the three peaks is then fit with a second order polynomial in phase and used to convert signals on that pixel to photon energy .
a given calibration solution is applied to observations near in time .
an analysis of the calibration solutions over time showed them to be consistent within 1 sigma between thermal cycles of the dewar or re - thresholding of the detectors .
a block diagram for the digital readout for arcons is shown in figure [ fig : readout ] and is described in detail in @xcite .
the digital readout for arcons utilizes eight reconfigurable open architecture computing hardware ( roach ) boards produced by the collaboration for astronomy signal processing and electronics research ( casper ) @xcite .
each roach board houses a virtex5 field programmable gate array ( fpga ) and a powerpc processor , and each board is connected to two digital - to - analog converters ( dacs ) , two analog - to - digital converters ( adcs ) , and a circuit board used for mixing signals , which is called the intermediate frequency ( if ) board .
the firmware on the fpga is used to control the output of the dac and to process the digitized signal from the adc .
the two outputs of the dac make up the in - phase and quadrature ( iq ) components of a complex signal used in a quadrature amplitude modulation strategy .
the dac is used to output a frequency comb that is mixed with a local oscillator ( lo ) frequency to move the comb into the frequency range of the mkid resonator frequencies .
this is then sent through the mkid array .
each frequency tone in the comb corresponds to one mkid resonator frequency .
the phase of the tone is set so the baseline phase of the same frequency read in the adc is approximately zero .
the amplitude of the tone is set according to the optimal readout power of the resonator addressed relative to the other resonators .
the complete frequency comb sent to the dac is retrieved by the fpga firmware from a look - up table ( lut ) stored in memory on the roach board .
the dacs and adcs are clocked at 512 mhz and the fpga is clocked at 256 mhz .
the frequency comb output from each dac can read out up to 256 resonators in 512 mhz of bandwidth , with 2 mhz spacing between tone frequencies .
the 2024 pixel array used in arcons is divided into two feedlines with 1012 pixels each .
four readout boards are connected to each feedline through power combiners and power splitters .
once the readout signal has passed through the mkids and has been mixed back down to baseband frequencies by the if board and digitized by the adc , the fpga firmware divides the signal into 256 frequency bins ( channels ) through a `` channelization '' process so that the contribution from each probe tone in the comb can be treated separately .
the firmware then searches for indications of photon absorption in the phase of each channel s signal .
these are seen as a sudden decrease in the phase followed by an exponential decay back to the baseline , with the characteristic time determined by the quasiparticle recombination time .
a template of this phase pulse shape was made for each pixel and used to create a customized optimal filters for that pixel .
the firmware applies the optimal filter to a channel s signal , then searches for negative peaks in the filtered phase that pass below a set threshold value .
each time the peak detection triggers , the firmware creates a 64-bit data packet containing the channel number ( so that the photon can be identified with a particular pixel ) , a timestamp , and values specifying the height of the phase pulse peak and the baseline phase level before the pulse to help remove any 1/f phase fluctuations .
a dead time of 100 @xmath1s is added after a trigger to keep pixels with incorrectly set trigger thresholds from filling up the buffers with false pulses .
a more elegant way of dealing with incorrect thresholds will be developed for future versions of the readout firmware .
the roach boards are synchronized using the 1 pps output of a meinberg gps170pex gps board and a stanford research fs725 rubidium 10 mhz frequency standard , which is needed for the timestamps to be recorded with 2 @xmath25 resolution .
the firmware outputs the data packets for the detected photons to two shared memory blocks on the roach .
a c program called pulseserver running on the powerpc located on each roach monitors the shared memory blocks . when they are filled to a certain point , pulseserver sends the data over a tcp connection to the data acquisition and control computer .
a c program called packetmaster , running on the data acquisition and control computer , receives the tcp data bundles from all eight roach boards .
it sifts through each photon packet and organizes them by pixel and second of arrival to write into an hdf data file .
packetmaster also creates images for the real - time display on the control computer .
the resulting hdf data file can later be processed with our software pipeline , detailed in section [ sec : pipeline ] .
arcons contains two independent microwave signal paths , which we refer to as feedline 1 and 2 .
each feedline reads out half of the array .
the signal path begins with a hermetic sma bulkhead connector that brings the comb of frequencies into the dewar .
a laser welded 0.087 `` stainless steel semi - rigid coaxial cable brings the signals from room temperature to 4 k , with a heat sink clamp cooling only the outer conductor at the 50 - 70 k stage . at 4 k
there is a 30 db attenuator , followed by an inner and outer dc block . after the dc block a 0.062 '' nbti / nbti ( outer conductor / center pin )
coax brings the signals to the 1 k stage , where the coax is interrupted with a heatsunk sma barrel connector .
a second nbti coax brings the signals to 100 mk . after passing through the array
, the signal goes through another nbti coax to 1 k , then a second to 4 k where it enters the input port of a hemt amplifier @xcite . after the hemt ,
another stainless coax brings the signals to a hermetic sma bulkhead connector and outside the dewar .
the arcons data reduction pipeline , though currently under development , is outlined here ( see figure [ fig : pipeline ] for a schematic ) .
the algorithms , written in python , will be made publicly available once the pipeline has become stable and reliable .
raw `` observation '' files consist of a set of tables , one for each pixel , with each row in the tables containing one second s worth of data packets , where each data packet represents a single photon event .
these files form the raw input to the pipeline .
the main steps in data reduction are as follows .
general outline of the arcons data reduction pipeline . ] barycenter photon arrival times : : the time of the start of each observation is stored in that observation hdf file s header .
the photon arrival timestamps recorded in the file are relative to this observation start time , so absolute arrival times at the location of the telescope are found by simply adding the recorded timestamps to the header time .
these timestamps can then be corrected to solar system barycentric times with the tempo2 pulsar timing package @xcite using a custom plug - in that allows it to treat photons individually .
the barycentered timestamps can then be used to compare the timing between signals recorded from different observatory locations without the effects of the earth s motion interfering .
cosmic ray cleaning : : the photon arrival times across all pixels are binned into histograms with overlapping @xmath26 bins , in order to search for multiple photon detection coincidences . where an interval is found with an unusually high number of coincident photon detections , that interval of @xmath20100 @xmath1s is flagged as a cosmic - ray contaminated interval for the entire chip .
the cosmic ray event rate is low enough that the integration time lost to this procedure is negligible .
wavelength calibration : : wavelength calibration files are created by uniformly exposing the detector array to the combined light of three lasers of known wavelength , as described in section [ sec : wavecal ] . for each photon event , the measured phase excursion is assumed to be a monotonically increasing function of photon energy .
a histogram of these phase shifts is created for each pixel . in a functional pixel
the histogram shows a superposition of three approximately normal distributions corresponding to the three laser wavelengths .
the locations of the maxima of these peaks are fit in order to find the phase shift corresponding to the three wavelengths . a second order polynomial fit to these three values , yields a smooth , continuous , analytic function , mapping phase shift to wavelength for each given pixel . flatfield calibration : : twilight sky exposures are used to normalize the response of each pixel with respect to each other .
the average twilight spectrum is computed by taking the median of the wavelength calibrated twilight spectrum for each pixel .
the ratio of this average spectrum to an individual pixel s wavelength calibrated spectrum then gives a weighting as a function of wavelength for that pixel .
this weighting is used to normalize the response of different pixels when combining data from multiple pixels , effecting the traditional ccd `` flatfielding '' but at every wavelength .
spectral shape calibration : : once the pixel responses are normalized relative to one other , a final absolute calibration of the overall spectral response is performed by comparing the spectrum of an observed reference target against its known spectrum . this provides a single additional whole - array wavelength weighting function .
the combination of this absolute weighting and the per - pixel relative flatfielding weights corresponds to the overall qe response of each pixel as a function of wavelength . applied together ,
the weights yield a fully calibrated spectrum for each pixel .
exposure time masking : : high levels of photon illumination induce a number of the pixels in the array to go `` hot '' from time to time , registering false counts and reporting artificially high count rates .
this is likely due to an interaction of the mkid with free electrons in the silicon wafer , which should be eliminated in new mkid designs .
this pixel behavior is detected by comparing the flux distribution in the neighborhood of each pixel to that of the expected psf ( similar to the algorithm used in the xmm - newton pipeline ) .
in each one - second time step , the ratio of the counts in each pixel to the median counts in a small surrounding box ( typically @xmath27 pixels ) is compared to the same ratio for a 2-dimensional gaussian function with fwhm matching that of the expected psf ( specified manually at the time of the data reduction ) . if , after accounting for photon shot noise errors , the measured ratio is significantly greater ( typically @xmath28 ) then the source of the counts can not be astrophysical , and the pixel is flagged as hot for the given timestep .
some pixels are also known to show anomalously _ low _ count rates ( `` cold '' pixels ) , or no counts at all ( `` dead '' pixels ) . in a similar way , therefore , if the count rate for a pixel is significantly lower than the median for the surrounding box , the pixel can be flagged as `` cold '' .
these intervals are then concatenated and tracked to maintain consistency in the total effective exposure time attributable to each pixel . since the number of bad pixels is relatively small compared to the total number of pixels , and the time intervals over which such unwanted behavior occur are small , the time masks are most efficiently stored as per - pixel lists of bad intervals , also in hdf file format .
this process is applicable to any array observation , including the laser wavelength calibrations and the twilight flats .
ra / dec determination : : an astrometric solution is found for the data at regular intervals ( typically 30s ) .
the centroid location of the target source in pixel coordinates is found using the python language `` pyguide '' centroiding algorithm , which calculates the point of minimum radial asymmetry for a source in a time integrated image .
the known sky coordinates of the target can then be assigned to that pixel location .
since the instrument location at the coud focus is fixed with respect to the telescope motion , the field of view also rotates at the sidereal rate .
the hour angle of the target therefore relates directly to the orientation of the field . in combination with a pre - calibrated
fixed rotational offset for the instrument , a right ascension and declination can be determined for every pixel at each time step .
in addition , atmospheric refraction spreads light of different frequencies onto different parts of the array .
this phenomenon is well understood @xcite and can be easily removed , allowing coordinates to be assigned for every photon .
tip / tilt correction : : for sources with sufficient count rates it should be possible to perform the above centroiding procedure on very short timescales , comparable with the timescale of atmospheric seeing variation ( @xmath29 ) .
alternatively , for non - point - source objects , image cross - correlation can be performed to determine offset corrections to the data on the same timescales .
hence it should be possible in principle to apply _ ex post facto _
first - order tip - tilt seeing corrections , purely in software , to improve the final recovered psf .
after these steps have been performed , processed photon lists will form the output data product : calibrated lists of photon events , with each photon having an absolute timestamp , ra and dec location , wavelength , and any associated flags for each photon , along with associated time - mask lists .
further analysis depends on the science goals and the desired end product . for imaging and spectroscopic analysis ,
image cubes are constructed by creating images integrated over different wavelength ranges and over the desired time intervals .
background subtraction is performed as in traditional ccd data reduction ( for example , by subtracting a constant or slowly varying two - dimensional polynomial fit to the image after masking bright sources ) .
spectra are then obtained by integrating over the desired pixels in the images . alternatively ,
science involving timing analysis ( e.g. , pulsar and eclipsing binary observations ) requires a non - discrete approach in the time - axis .
we integrate first over selected pixels or a selected area in sky - coordinates , and maintain the individual photon timing information , avoiding binning into larger time intervals . in the case of pulsar or eclipsing binary observations , for example ,
this allows for folding of the data on the known period .
even with the current engineering - grade mkid arrays , the extra mirror bounces and path length of the coud focus , and non - optimal optical design and anti - reflection coatings , arcons is one of the most powerful tools for narrow field of view astrophysics ever developed .
it has now been proven in the field over the course of 24 observing nights at 5-m class telescopes . in this section
we explore in detail the performance of the instrument .
typically over 90% of the resonators show up in frequency sweeps , but due to the variations in the tin gap there are a significant number of collisions , reducing usable pixels , as shown in figure [ fig : yield ] . after cutting out collisions and pixels with especially high or low quality factor , usually about 70% of our pixels are usable
. a more uniform film , such as the ones made at nist with ti / tin multilayers @xcite or with atomic layer deposition ( ald ) , could significantly reduce the number of collisions , dramatically improving yield .
the typical spectral resolution for the mkids used in arcons is r@xmath2010 at 4000 , declining linearly with increasing wavelength as expected .
the median spectral resolution is degraded to r@xmath208 with our two channel analog electronics ( see figure [ fig : res ] ) with the full arcons digital readout due to limitations in the number of coefficients ( taps ) in the programmable optimal filter in our digital electronics .
this issue is being addressed by upgrading our electronics from roach to roach2 , which will allow for significantly more taps .
cts / sec .
the green curve shows the projected energy resolution we expect to recover at the nominal sky count rate of 100 cts / sec by scaling the black data by the expected degradation in spectral resolution with count rate , the red curve in figure [ fig : lin ] .
the other feedline is very similar . ]
the achieved spectral resolution is drastically below the theoretical limit for a 100 mk operating temperature of r=100 at 4000 @xcite .
we believe the current mkids are limited by positional effects within the mkid inductor due to local variations in the tin superconducting gap . fixing this should roughly double the spectral resolution .
after this issue is addressed , lower noise microwave amplifiers @xcite and new mkid designs should continue to boost the resolution towards the theoretical limit .
as shown in figure [ fig : lin ] as a red line ( labeled on the right y - axis ) , there is also a pronounced degradation of the spectral resolution with increasing count rate .
this is caused by two effects .
first , the baseline value that the firmware tracks to reduce 1/f noise gets noisy from averaging many counts .
second , pulses begin overlapping , degrading resolution
. improvements in the firmware baseline determination algorithm should improve this degradation significantly . at palomar ,
the arcons sky background rate is 50100 photons / sec , so only objects significantly brighter than sky will suffer from severe degradation of the spectral resolution .
the mkids in arcons absorb light directly in the tin film that comprises the inductor .
this tin has an intrinsic absorption of roughly 70% at 4000 , and 30% at 1 .
the total system throughput is reduced by the cumulative effect of all the intervening optical materials and surfaces ; the atmosphere , the telescope m1m4 mirrors ( the coud m3 at palomar was degraded due to @xmath30 years since realuminization ) , the collimating lens , the asahi supercold filter at 4 k , the 4 optical elements at 4k , the 100 mk glass with ito filter , and the microlens fill factor and alignment with the inductors .
a model of this entire system , based on a simple atmospheric model , clean al reflectivity ( normalized to 86% at 670 nm ) , manufacturer s specs for glasses , ar coatings , and the supercold filter , and measurements of the absorptivity of raw tin films , is shown as a red solid line in figure [ fig : sens ] .
the efficiency of just arcons , neglecting the atmosphere and telescope , is shown as the red dashed line .
the theoretical model of arcons can be directly compared with laboratory measurements of arcons s qe . these tests were done with a qe testbed consisting of a monochromater and light source , integrating sphere , and a rotatable fold mirror to direct the beam to a calibrated photodiode or to the arcons optical input .
the results of these lab tests are shown as a black dashed line in figure [ fig : sens ] .
the roughly factor of two discrepancy between the theoretical model and the laboratory measurements is primarily due to a known misalignment of the microlens height above the focal plane , increasing the spot size on the focal plane so it is larger than the inductor , especially at red wavelengths .
the actual system throughput is calculated by doing photometry on a standard star and comparing our results to the object s known spectrum , shown as the solid black line in figure [ fig : sens ] .
this curve is about a factor of two below the laboratory qe measurements modified by the expected atmospheric transmission and telescope throughout .
the cause of this discrepancy is unknown , but could include the unknown reflectivity of the coud m3 flat or angular misalignment of arcons with respect to the telescope input .
the mkids used is arcons have inductors with a preferred orientation of the wires , which could potentially lead to a quantum efficiency that depends on polarization angle .
lab measurements were performed on arcons to check whether the mkid arrays exhibited any response to linearly polarized light at different angles .
no polarization response was seen to the 5% level . by going to cassegrain focus , improving our anti - reflection coatings , aligning the microlens well , and simplifying our optics
we should be able to boost our total system throughput in future mkid cameras by a factor of 10 .
using mkids with black absorbers could eventually yield instruments with total system efficiencies greater than 60% .
the 100 @xmath1s dead time associated with triggering ( section [ sec : readout ] ) causes some pulses to be missed at high count rates , as shown in figure [ fig : lin ] .
this effect can be corrected for analytically by calculating the number of photons that are expected to arrive within a dead time interval and hence not be counted , @xmath31 , where @xmath32 is the corrected count rate , @xmath33 is the measured count rate , and @xmath34 is the dead time , as shown in as a green line in figure [ fig : lin ] .
the slightly longer effective dead time of 120 @xmath1s used in figure [ fig : lin ] results from the interaction of the firmware dead time and the pulse optimal filters .
the firmware of arcons is currently being rewritten to remove the dead time and better handle closely spaced photons , which should significantly improve the linearity and spectral resolution at high count rates .
imaging and spectroscopic performance of arcons is consistent with expectations from the beam map and spectral resolution data .
the plate scale was measured to be 0.435``/pixel , matching well with our zemax design . during the run we experienced one night with seeing measured at @xmath31.0 '' by the palomar
seeing monitor , and fitting a psf from this night shows that we recover this seeing value , indicating that our optics are not limiting our imaging performance .
extremely high quality imaging and spectra will require more of our software pipeline to be in place , but preliminary analyses are presented in figures [ fig : arp ] and [ fig : spectra ] .
these images and spectra begin to show the capabilities of this unique instrument .
arcons timing performance was measured by simultaneously observing the crab pulsar with arcons and the green bank telescope ( gbt ) .
the data shown in figure [ fig : crab ] shows that arcons sees an optical pulse leading the radio pulse by about 150 @xmath1s , which is consistent with previous observations @xcite .
measurement of the arrival time of a single photon should be accurate @xmath35s .
currently , there is a delay of 43 @xmath1s from the time a photon is absorbed by the mkid until the photon packet is ready to be sent over the network , opening up the possibility of extremely fast effective frame rates for certain applications , like wavefront sensing and speckle control in a coronagraph .
arcons is a unique , read - noise free photon counting ifs . with a 2024 mkid pixel array covering a 20``@xmath020 '' field of view , arcons is the world s largest ( and only active ) ltd - based instrument in the optical through near - ir .
it is uniquely powerful for observations of rapidly variable sources .
the arcons instrument has performed astronomical observations , and the first science papers using arcons data are currently being written .
arcons and its successors will continue to improve in pixel count and yield , spectral resolution , system throughput , and detector quantum efficiency . the mkid technology it uses is extremely scalable , allowing arrays approaching megapixels within a decade .
mkid - based instruments like arcons will start to become a standard part of the uvoir observer s toolkit in the years to come .
the mkid detectors used in this work were developed under nasa grant nnx11ad55 g . s.r .
meeker was supported by a nasa office of the chief technologist s space technology research fellowship , nasa grant nnx11an29h .
this work was partially supported by the keck institute for space studies .
fermilab is operated by fermi research alliance , llc under contract no .
de - ac02 - 07ch11359 with the united states department of energy .
the authors would also like to thank shri kulkarni , director of the caltech optical observatories , and jason prochaska , associate director of lick observatory , for facilitating this project , as well as the excellent staffs of the palomar and lick observatories for their assistance in getting arcons working .
jennifer milburn s help with the guide camera software was invaluable .
this project also greatly benefitted from the support of mike werner , paul goldsmith , and jonas zmuidzinas at jpl . | we present the design , construction , and commissioning results of arcons , the * * ar**ray * * c**amera for * * o**ptical to * * n**ear - ir * * s**pectrophotometry .
arcons is the first ground - based instrument in the optical through near - ir wavelength range based on microwave kinetic inductance detectors ( mkids ) .
mkids are revolutionary cryogenic detectors , capable of detecting single photons and measuring their energy without filters or gratings , similar to an x - ray microcalorimeter .
mkids are nearly ideal , noiseless photon detectors , as they do not suffer from read noise or dark current and have nearly perfect cosmic ray rejection .
arcons is an integral field spectrograph ( ifs ) containing a lens - coupled 2024 pixel mkid array yielding a 20``@xmath020 '' field of view , and has been deployed on the palomar 200 `` and lick 120 '' telescopes for 24 nights of observing .
we present initial results showing that arcons and its mkid arrays are now a fully operational and powerful tool for astronomical observations . |
as a warning to the reader , we want to stress from the beginning that this paper is mostly a formal exercise : to understand how the bayesian analysis of a mixture model unravels and automatically exploits the missing data structure of the model is crucial for grasping the details of simulation methods ( not covered in this paper , see , e.g. , @xcite ) that take full advantage of the missing structures .
it also allows for a comparison between exact and approximate techniques when the former are available . while the relevant references are pointed out in due time , we note here that our paper builds upon the foundational paper of @xcite .
we thus assume that a sample @xmath0 from the mixture model @xmath1 is available , where @xmath2 denotes the scalar product between the vectors @xmath3 and @xmath4 .
we are selecting on purpose the natural representation of an exponential family ( see , e.g. * ? ? ? * chapter 3 ) , in order to facilitate the subsequent derivation of the posterior distribution .
when the components of the mixture are poisson @xmath5 distributions , if we define @xmath6 , the poisson distribution indeed is written as a natural exponential family : @xmath7 for a mixture of multinomial distributions @xmath8 , the natural representation is given by @xmath9 and the overall ( natural ) parameter is thus @xmath10 . in the normal @xmath11 case ,
the derivation is more delicate when both parameters are unknown since @xmath12 in this particular setting , the natural parameterisation is in @xmath13 while the statistic @xmath14 is two - dimensional .
the moment cumulant function is then @xmath15 .
as described in the standard literature on mixture estimation @xcite , the missing variable decomposition of a mixture likelihood associates each observation in the sample with one of the @xmath16 components of the mixture , i.e. @xmath17 given the component allocations @xmath18 , we end up with a cluster of ( sub)samples from different distributions from the same exponential family .
priors customarily used for the analysis of these exponential families can therefore be extended to the mixtures as well .
while conjugate priors do not formally exist for mixtures of exponential families , we will define _ locally conjugate priors _ as priors that are conjugate for the completed distribution , that is , for the likelihood associated with both the observations and the missing data @xmath18 .
this amounts to taking regular conjugate priors for the parameters of the different components and a conjugate dirichlet prior on the weights of the mixture , @xmath19 when we consider the complete likelihood @xmath20 \\ & = \prod_{j=1}^k p_j^{n_j } \exp\left [ \theta_j\cdot \sum_{z_i = j } r(x_i ) - n_j\psi(\theta_j ) \right ] \\ & = \prod_{j=1}^k p_j^{n_j } \exp\left [ \theta_j\cdot s_j - n_j\psi(\theta_j ) \right]\,,\end{aligned}\ ] ] it is easily seen that we remain within an exponential family since there exists a sufficient statistic with fixed dimension , @xmath21 .
if we use a dirichlet prior , @xmath22 on the vector of the weights @xmath23 defined on the simplex of @xmath24 , and ( generic ) conjugate priors on the @xmath25s , @xmath26 \propto \exp\left [ \theta_j\cdot s_{0j } - \lambda_j\psi(\theta_j ) \right ] \,,\ ] ] the posterior associated with the complete likelihood @xmath27 is then of the same family as the prior : @xmath28 \times p_j^{n_j } \exp\left [ \theta_j\cdot s_j - n_j\psi(\theta_j ) \right ] \\ & = \prod_{j=1}^k p_j^{\alpha_j+n_j-1}\,\exp\left [ \theta_j\cdot ( s_{0j}+s_j ) - ( \lambda_j+n_j)\psi(\theta_j ) \right]\,;\end{aligned}\ ] ] the parameters of the prior are transformed from @xmath29 to @xmath30 , from @xmath31 to @xmath32 and from @xmath33 into @xmath34 .
for instance , in the case of the poisson mixture , the conjugate priors are gamma @xmath35 , with corresponding posteriors ( for the complete likelihood ) , gamma @xmath36 distributions , in which @xmath37 denotes the sum of the observations in the @xmath38th group . for a mixture of multinomial distributions , @xmath39 ,
the conjugate priors are dirichlet @xmath40 distributions , with corresponding posteriors @xmath41 , @xmath42 denoting the number of observations from component @xmath38 in group @xmath43 @xmath44 , with @xmath45 . in the normal mixture case ,
the standard conjugate priors are products of normal and inverse gamma distributions , i.e. @xmath46 indeed , the corresponding posterior is @xmath47 and @xmath48 where @xmath49 is the sum of the observations allocated to component @xmath38 and @xmath50 is the sum of the squares of the differences from @xmath51 for the same group ( with the convention that @xmath52 when @xmath53 ) . these straightforward derivations do not correspond to the observed likelihood , but to the completed likelihood .
while this may be enough for some simulation methods like gibbs sampling ( see , e.g. * ? ? ?
* ; * ? ? ?
* ) , we need further developments for obtaining the true posterior distribution . if we now consider the observed likelihood , it is natural to expand this likelihood as a sum of completed likelihoods over all possible configurations of the partition space of allocations , that is , a sum over @xmath54 terms . except in the very few cases that are processed below , including poisson and multinomial mixtures ( see section [ ex : fullp ] ) , this sum does not simplify into a smaller number of terms because there exists no summary statistics . from a bayesian point of view
, the complexity of the model is therefore truly of magnitude @xmath55 .
the observed likelihood is thus @xmath56 ( with the dependence of @xmath57 upon @xmath18 omitted for notational purposes ) and the associated posterior is , up to a constant , @xmath58 where @xmath59 is the normalising constant missing in @xmath60 i.e. @xmath61 if @xmath62 is the normalising constant of @xmath63 , i.e. @xmath64 the posterior @xmath65 is therefore a mixture of conjugate posteriors where the parameters of the components as well as the weights can be computed in closed form !
the availability of the posterior does not mean that alternative estimates like map and mmap estimates can be computed easily
. however , this is a useful closed form result in the sense that moments can be computed exactly : for instance , if there is no label switching problem @xcite and , if the posterior mean is producing meaningful estimates , we have that @xmath66 = \sum_{{\mathbf{z}}}\ , \omega({\mathbf{z}})\ , \frac{s_{0j}+s_j}{n_j+\lambda_j\,,}\ ] ] since , for each allocation vector @xmath18 , we are in an exponential family set - up where the posterior mean of the expectation @xmath67 of @xmath4 is available in closed form .
( obviously , the posterior mean only makes sense as an estimate for very discriminative priors ; see @xcite . )
similarly , estimates of the weights @xmath68 are given by @xmath69 = \sum_{{\mathbf{z}}}\ , \omega({\mathbf{z}})\ , \frac{n_j+\alpha_j}{n+\alpha_\cdot}\,,\ ] ] where @xmath70 .
therefore , the only computational effort required is the summation over all partitions .
this decomposition further allows for a closed form expression of the marginal distributions of the various parameters of the mixture .
for instance , the ( marginal ) posterior distribution of @xmath71 is given by @xmath72}{k(s_{0j}+s_j , n_j+\lambda_j)}\,.\ ] ] ( note that , when the hyperparameters @xmath29 , @xmath31 , and @xmath73 are independent of @xmath38 , this posterior distribution is independent of @xmath38 . ) similarly , the posterior distribution of the vector @xmath23 is equal to @xmath74 if @xmath16 is small and @xmath75 is large , and when all hyperparameters are equal , the posterior should then have @xmath16 spikes or peaks , due to the label switching / lack of identifiability phenomenon .
we will now proceed through standard examples . in the case of a two component poisson mixture , @xmath76
let us assume a uniform prior on @xmath77 ( i.e. @xmath78 ) and exponential priors @xmath79 and @xmath80 on @xmath81 and @xmath82 , respectively . (
the scales are chosen to be fairly different for the purpose of illustration . in a realistic setting
, it would be sensible either to set those scales in terms of the scale of the problem , if known , or to estimate the global scale following the procedure of @xcite . )
the normalising constant is then equal to @xmath83 { \,\text{d}\,}\theta\\ & = \int_0^\infty \lambda_j^{\xi-1}\,\exp ( -\delta\lambda_j)\,{\,\text{d}\,}\lambda_j \\ & = \delta^{-\xi}\,\gamma(\xi)\,,\end{aligned}\ ] ] with @xmath84 and @xmath85 , and the corresponding posterior is ( up to the normalisation of the weights ) @xmath86 with @xmath87 corresponding to a beta @xmath88 distribution on @xmath77 and to a gamma @xmath89 distribution on @xmath33 @xmath90 .
an important feature of this example is that the sum does not need to involve all of the @xmath91 terms , simply because the individual terms in the previous sum factorise in @xmath92 , which then acts like a local sufficient statistic . since @xmath93 and @xmath94 , the posterior only requires as many distinct terms as there are distinct values of the pair @xmath95 in the completed sample .
for instance , if the sample is @xmath96 , the distinct values of the pair @xmath95 are @xmath97 .
there are therefore @xmath98 distinct terms in the posterior , rather than @xmath99 . the problem of computing the number ( or cardinality ) @xmath100 of terms in the @xmath54 sum with the same statistic
@xmath95 has been tackled by @xcite in that he proposes a recursive formula for computing @xmath101 in an efficient way , as expressed below for a @xmath16 component mixture : * @xcite * if @xmath102 denotes the vector of length @xmath16 made up of zeros everywhere except at component @xmath38 where it is equal to one , if @xmath103 then @xmath104 therefore , once the @xmath105 are all computed , the posterior can be written as @xmath106 \ , \pi(\theta,\mathbf{p}|{\mathbf{x}},n_1,s_1)\,,\ ] ] up to a constant , since the complete likelihood posterior only depends on the sufficient statistic @xmath95 .
now , the closed - form expression allows for a straightforward representation of the marginals .
for instance , the marginal in @xmath81 is given by @xmath107 up to a constant , while the marginal in @xmath82 is @xmath108 again up to a constant , and the marginal in @xmath77 is @xmath109 still up to a constant , if @xmath110 denotes the sum of all observations .
as pointed out above , another interesting outcome of this closed - form representation is that marginal likelihoods ( or evidences ) can also be computed in closed form .
the marginal distribution of @xmath111 is directly related to the unormalised weights @xmath112 in that @xmath113 up to the product of factorials @xmath114 ( but this is irrelevant in the computation of the bayes factor ) . in practice
, the derivation of the cardinalities @xmath100 can be done recursively as in @xcite : include each observation @xmath115 by updating all the @xmath116s in both @xmath117 and @xmath118 , and then check for duplicates . below is a nave r implementation ( for reasonable efficiency , the algorithm should be programmed in a faster language like c. ) , where ` ncomp ` denotes the number of components : .... # matrix of sufficient statistics , last column is number of occurrences cardin = matrix(0,ncol=2*ncomp+1,nrow = ncomp ) # initialisation for ( i in 1:ncomp ) cardin[i,((2*i)-1):(2*i)]=c(1,dat[1 ] ) cardin[,2*ncomp+1]=1 # update for ( i in 2:length(dat ) ) { ncard = dim(cardin)[1 ] update = matrix(t(cardin),ncol=2*ncomp+1,nrow = ncomp*ncard , byrow = t ) for ( j in 0:(ncomp-1 ) ) { update[j*ncard+(1:ncard),(2*j)+1]= update[j*ncard+(1:ncard),(2*j)+1]+1 update[j*ncard+(1:ncard),(2*j)+2]= update[j*ncard+(1:ncard),(2*j)+2]+dat[i ] } update = update[do.call(order , data.frame(update ) ) , ] nu = dim(update)[1 ] # changepoints jj = c(1,(2:nu)[apply(abs(update[2:nu,1:(2*ncomp)]- update[1:(nu-1),1:(2*ncomp)]),1,sum)>0 ] ) # duplicates or rather ncomplicates !
duplicates=(1:nu)[-jj ] if ( length(duplicates)>0 ) { for ( dife in 1:(ncomp-1 ) ) { ji = jj[jj+dife<=nu ] ii = ji[apply(abs(update[ji+dife,1:(2*ncomp)]- update[ji,1:(2*ncomp)]),1,sum)==0 ] if ( length(ii)>0 ) update[ii,(2*ncomp)+1]=update[ii,(2*ncomp)+1]+ update[ii+dife,(2*ncomp)+1 ] } update = update[-duplicates , ] } cardin = update } .... at the end of this program , all non - empty realisations of the sufficient @xmath95 are available in the two first columns of ` cardin ` , while the corresponding @xmath100 is provided by the last column .
once the @xmath100 s are available , the corresponding weights can be added as the last column of ` cardin ` , i.e. .... w = log(cardin[,2*ncomp+1])+apply(lfactorial(cardin[,2*(1:ncomp)-1]),1,sum)+ apply(lfactorial(cardin[,2*(1:ncomp)]),1,sum)- apply(log(xi[1:ncomp]+cardin[,2*(1:ncomp)-1 ] ) * ( cardin[,2*(1:ncomp)]+1),1,sum)- sum(lfactorial(dat ) ) w = exp(w - max(w ) ) cardin = cbind(cardin , w ) .... where ` xi[j ] ` denotes @xmath31 .
the marginal posterior on @xmath81 can then be plotted via .... marlam = function(lam , comp=1 ) { sum(cardin[,2*(ncomp+1)]*dgamma(lam , shape = cardin[,2*comp]+1 , rate = cardin[,2*comp-1]+xi[comp]))/sum(cardin[,2*(ncomp+1 ) ] ) } lalam = seq(.01,1.2*max(dat),le=100 ) mamar = apply(as.matrix(lalam),1,marlam , comp=1 ) plot(lalam , mamar , type="l",xlab = expression(mu[1]),ylab="",lwd=2 ) .... while the marginal posterior on @xmath77 is given through .... marp = function(p , comp=1 ) { sum(cardin[,2*(ncomp+1)]*dbeta(p , shape1=cardin[,2*comp-1]+1 , shape2=length(dat)-cardin[,2*comp-1]+1))/sum(cardin[,2*(ncomp+1 ) ] ) } pepe = seq(.01,.99,le=99 ) papar = apply(as.matrix(pepe),1,marp ) plot(pepe , papar , type="l",xlab="p",ylab="",lwd=2 ) .... now , even with this considerable reduction in the complexity of the posterior distribution ( to be compared with @xmath54 ) , the number of terms in the posterior still grows very fast both with @xmath75 and with the number of components @xmath16 , as shown through a few simulated examples in table [ tab : explose ] .
( the missing items in the table simply took too much time or too much memory on the local mainframe when using our ` r ` program .
@xcite used a specific ` c ` program to overcome this difficulty with larger sample sizes . )
the computational pressure also increases with the range of the data ; that is , for a given value of @xmath119 , the number of rows in ` cardin ` is much larger when the observations are larger , as shown for instance in the first three rows of table [ tab : explose ] : a simulated poisson @xmath120 sample of size @xmath121 is primarily made up of zeros when @xmath122 but mostly takes different values when @xmath123 .
the impact on the number of sufficient statistics can be easily assessed when @xmath124 .
( note that the simulated dataset corresponding to @xmath125 in table [ tab : explose ] corresponds to a sample only made up of zeros , which explains the @xmath126 values of the sufficient statistic @xmath127 when @xmath128 . )
l ccc @xmath129 & @xmath128 & @xmath130 & @xmath124 + @xmath131 & 11 & 66 & 286 + @xmath132 & 52 & 885 & 8160 + @xmath133 & 166 & 7077 & 120,908 + @xmath134 & 57 & 231 & 1771 + @xmath135 & 260 & 20,607 & 566,512 + @xmath136 & 565 & 100,713 & + @xmath137 & 87 & 4060 & 81,000 + @xmath138 & 520 & 82,758 & + @xmath139 & 1413 & 637,020 & + @xmath140 & 216 & 13,986 & + @xmath141 & 789 & 271,296 & + @xmath142 & 2627 & & + an interesting comment one can make about this decomposition of the posterior distribution is that it may happen that , as already noted in @xcite , a small number of values of the local sufficient statistic @xmath95 carry most of the posterior weight .
table [ tab : cumuweit ] provides some occurrences of this feature , as for instance in the case @xmath143 .
l ccc @xmath129 & @xmath128 & @xmath130 & @xmath124 + @xmath132 & 20/44 & 209/675 & 1219/5760 + @xmath133 & 58/126 & 1292/4641 & 13,247/78,060 + @xmath134 & 38/40 & 346/630 & 1766/6160 + @xmath135 & 160/196 & 4533/12,819 & 80,925/419,824 + @xmath144 & 99/314 & 5597/28,206 & + @xmath145 & 21/625 & 13,981/117,579 & + @xmath146 & 50/829 & 62,144/211,197 & + @xmath136 & 1/580 & 259/103,998 & + @xmath137 & 198/466 & 20,854/70,194 & 30,052/44,950 + @xmath138 & 202/512 & 18,048/80,470 & + @xmath147 & 1/1079 & 58,820/366,684 & + we now turn to a minnow dataset made of @xmath148 observations , for which we need a minimal description .
as seen in figure [ fig : topminnow1 ] , the datapoints take large values , which is a drawback from a computational point of view since the number of statistics to be registered is much larger than when all datapoints are small .
for this reason , we can only process the mixture model with @xmath128 components . _
( top right ) _ marginal posterior distribution of @xmath82 _ ( bottom left ) _ marginal posterior distribution of @xmath77 _ ( bottom right ) _ histogram of the minnow dataset .
( the prior parameters are @xmath149 and @xmath150 to remain compatible with the data range . ) ] if we instead use a completely symmetric prior with identical hyperparameters for @xmath81 and @xmath82 , the output of the algorithm is then also symmetric in both components , as shown by figure [ fig : topminnow2 ] .
the modes of the marginals of @xmath81 and @xmath82 remain the same , nonetheless . for a symmetric prior with hyperparameter @xmath149 . ]
the case of a multinomial mixture can be dealt with similarly : if we have @xmath75 observations @xmath151 from the mixture @xmath152 where @xmath153 and @xmath154 , the conjugate priors on the @xmath155s are dirichlet distributions @xmath156 , @xmath157 and we use once again the uniform prior on @xmath77 .
( a default choice for the @xmath158 s is @xmath159 . )
note that the @xmath160s may differ from observation to observation , since they are irrelevant for the posterior distribution : given a partition @xmath18 of the sample , the complete posterior is indeed @xmath161 up to a normalising constant that does not depend on @xmath18 .
more generally , if we consider a mixture with @xmath162 components , @xmath163 the complete posterior is also directly available , as @xmath164 once more up to a normalising constant .
the corresponding normalising constant of the dirichlet distribution being @xmath165 it produces the overall weight of a given partition @xmath18 as @xmath166 where @xmath167 is the number of observations allocated to component @xmath168 , @xmath169 is the sum of the @xmath170s for the observations @xmath171 allocated to component @xmath168 and @xmath172 given that the posterior distribution only depends on those sufficient " statistics @xmath173 and @xmath167 , the same factorisation as in the poisson case applies , namely that we simply need to count the number of occurrences of a particular local sufficient statistic @xmath174 .
the book - keeping algorithm of @xcite applies in this setting as well .
what follows is a nave r program translating the above : .... em = dim(dat)[2 ] emp = em+1 empcomp = emp*ncomp # matrix of sufficient statistics : # last column is number of occurrences # each series of ( em+1 ) columns contains , first , number of allocations # and , last , sum of multinomial observations cardin = matrix(0,ncol = empcomp+1,nrow = ncomp ) .... therefore , the @xmath175th column of ` cardin ` contains the sum of the @xmath160s for the @xmath38 s allocated to the first component . .... # initialisation for ( i in 1:ncomp ) cardin[i , emp*(i-1)+(1:emp)]=c(1,dat[1 , ] ) cardin[,empcomp+1]=1 # update for ( i in 2:dim(dat)[1 ] ) { ncard = dim(cardin)[1 ] update = matrix(t(cardin),ncol = empcomp+1,nrow = ncomp*ncard , byrow = t ) for ( j in 0:(ncomp-1 ) ) { indi = j*ncard+(1:ncard ) empj = emp*j update[indi , empj+1]=update[indi , empj+1]+1 update[indi , empj+(2:emp)]=t(t(update[indi , empj+(2:emp)])+dat[i , ] ) } update = update[do.call(order , data.frame(update ) ) , ] nu = dim(update)[1 ] # changepoints jj = c(1,(2:nu)[apply(abs(update[2:nu,1:empcomp]-update[1:(nu-1 ) , 1:empcomp]),1,sum)>0 ] ) # duplicates or rather ncomplicates !
duplicates=(1:nu)[-jj ] if ( length(duplicates)>0 ) { for ( dife in 1:(ncomp-1 ) ) { ji = jj[jj+dife<=nu ] ii = ji[apply(abs(update[ji+dife,1:empcomp]- update[ji,1:empcomp]),1,sum)==0 ] if ( length(ii)>0 ) update[ii , empcomp+1]=update[ii , empcomp+1]+ update[ii+dife , empcomp+1 ] } update = update[-duplicates , ] } cardin = update # print(sum(cardin[,2*ncomp+1])-ncomp^i ) } .... where ` dat ` is now a matrix with @xmath16 columns . the computation of the number of replicates of a given sufficient statistic @xmath176 @xmath177 , is then provided by the last column of the matrix ` cardin ` .
the overall weight is then computed as the product of @xmath177 with the normalising constant : .... olsums = matrix(0,ncol = ncomp , nrow = dim(update)[1 ] ) for ( y in 1:ncomp ) colsums[,y]=apply(update[,(y-1)*emp+(2:emp)],1,sum ) w = log(cardin[,empcomp+1])+ apply(lfactorial(cardin[,emp*(0:(ncomp-1))+1]),1,sum)+ apply(lfactorial(cardin [ , ( 1:empcomp)[-1-emp*(0:(ncomp-1))]]-.5),1,sum)- apply(lfactorial(colsums)+em*.5 - 1,1,sum)- sum(lfactorial(dat ) ) w = exp(w - max(w ) ) cardin = cbind(cardin , w ) .... as shown in table [ tab : multi ] , once again , the reduction in the number of cases to be considered is enormous .
l cccc @xmath178 & @xmath179 & @xmath180 & @xmath181 & @xmath182 + @xmath183 & @xmath184 & @xmath185 & @xmath186 & + @xmath187 & @xmath188 & @xmath189 & @xmath190 & + @xmath191 & @xmath192 & @xmath193 & @xmath194 & + @xmath195 & @xmath196 & @xmath197 & @xmath198 & @xmath199 + @xmath200 & @xmath201 & @xmath202 & @xmath203 & + @xmath204 & @xmath205 & @xmath206 & @xmath207 & + @xmath208 & @xmath209 & @xmath210 & @xmath211 & + @xmath212 & @xmath213 & @xmath214 & @xmath215 & @xmath216 + @xmath217 & @xmath218 & @xmath219 & @xmath220 & + @xmath221 & @xmath222 & @xmath223 & & + @xmath224 & @xmath225 & @xmath226 & & + _ ( missing terms are due to excessive computational or storage requirements . ) _ for a normal mixture ,
the number of truly different terms in the posterior distribution is much larger than in the previous ( discrete ) cases , in the sense that only permutations of the members of a given partition within each term of the partition provide the same local sufficient statistics .
therefore , the number of observations that can be handled in an exact analysis is necessarily extremely limited .
as mentioned in section [ sub : locconj ] , the locally conjugate priors for normal mixtures are products of normal @xmath227 by inverse gamma @xmath228 distributions .
for instance , in the case of a two - component normal mixture , @xmath229 we can pick @xmath230 , @xmath231 , @xmath232 , if a difference of one between both means is considered likely ( meaning of course that the data are previously scaled ) and if @xmath233 is the prior assumption on the variance ( possibly deduced from the range of the sample ) . obviously , the choice of a gamma distribution with @xmath234 degrees of freedom is open to discussion , as it is not without consequences on the posterior distribution .
the normalising constant of the prior distribution is ( up to a true constant ) @xmath235 indeed , the corresponding posterior is @xmath236\ ] ] and @xmath237\,.\ ] ] the number of different sufficient statistics @xmath238 is thus related to the number of different partitions of the dataset into at most @xmath16 groups .
this is related to the bell number @xcite , which grows extremely fast .
we therefore do not pursue the example of the normal mixture any further for lack of practical purpose .
this paper is a chapter of the book _ mixtures : estimation and applications _ , edited by the authors jointly with mike titterington and following the icms workshop on the same topic that took place in edinburgh , march 03 - 05 , 2010 .
the authors are deeply grateful to the staff at icms for the organisation of the workshop , to the funding bodies ( epsrc , lms , edinburgh mathematical society , glasgow mathematical journal trust , and royal statistical society ) for supporting this workshop , and to the participants in the workshop for their innovative and exciting contributions .
lee k , marin jm , mengersen k and robert c 2009 bayesian inference on mixtures of distributions in _ perspectives in mathematical sciences i : probability and statistics _
sastry nn , delampady m and rajeev b ) , pp .
world scientific singapore .
mengersen k and robert c 1996 testing for mixtures : a bayesian entropic approach ( with discussion ) in _ bayesian statistics 5 _ ( ed berger j , bernardo j , dawid a , lindley d and smith a ) , pp . 255276 .
oxford university press . | in this paper , we show how a complete and exact bayesian analysis of a parametric mixture model is possible in some cases when components of the mixture are taken from exponential families and when conjugate priors are used .
this restricted set - up allows us to show the relevance of the bayesian approach as well as to exhibit the limitations of a complete analysis , namely that it is impossible to conduct this analysis when the sample size is too large , when the data are not from an exponential family , or when priors that are more complex than conjugate priors are used .
* keywords : * bayesian inference , conjugate prior , exponential family , poisson mixture , binomial mixture , normal mixture . |
in 2002 , the selex collaboration reported the first observation of a candidate for a double charmed baryon , decaying as @xmath6@xcite .
the state had a mass of 3519 @xmath4 2 , and its observed width was consistent with experimental resolution , less than 5 @xmath7 .
the final state contained a charmed hadron , a baryon , and negative strangeness ( and @xmath8 ) , consistent with the cabibbo - allowed decay of a configuration . in order to confirm the interpretation of this state as a double charm baryon ,
it is essential to observe the same state in some other way .
other experiments with large charm baryon samples , e.g. , the focus and e-791 fixed target charm experiments at fermilab or the b - factories , have not confirmed the double charm signal .
this is consistent with the selex results .
the report in ref .
@xcite emphasized that this new state was produced by the baryon beams ( @xmath9 , proton ) in selex , but not by the @xmath10 beam .
it also noted that the apparent lifetime of the state was significantly shorter than that of the @xmath11 , which was not expected in a model calculation based on heavy quark effective theory @xcite . another way to confirm the
is to observe it in a different decay mode that also involves a final state with baryon number and charm ( not anti - charm ) .
one such mode , involving only stable charged particles , is the channel . observing a mass peak near the peak at 3519 reported in ref .
@xcite in a channel combining a proton with a @xmath12 pair but not a @xmath13 pair would confirm the existence of the state .
here we report the first observation of .
the selex experiment used the fermilab charged hyperon beam at 600 gev to produce charm particles in a set of thin foil targets of cu or diamond .
the negative beam composition was about 50% @xmath9 , 50% @xmath10 .
the positive beam was 90% protons .
a beam transition radiation detector identified each beam particle as meson or baryon with zero overlap .
the three - stage magnetic spectrometer is shown elsewhere @xcite .
the most important features are the high - precision , highly redundant , vertex detector that provided an average proper time resolution of 20 for the charm decays , a 10 m long ring - imaging cerenkov ( rich ) detector that separated @xmath14 from k up to 165 gev / c @xcite , and a high - resolution tracking system that had momentum resolution of @xmath15 for a proton .
the experiment selected charm candidate events using an online secondary vertex algorithm .
a scintillator trigger demanded an inelastic collision with at least four charged tracks in the interaction scintillators and at least two hits in the positive particle hodoscope after the second analyzing magnet .
event selection in the online filter required full track reconstruction for measured fast tracks ( @xmath16 ) .
these tracks were extrapolated back into the vertex silicon planes and linked to silicon hits .
the beam track was measured in upstream silicon detectors .
a three - dimensional vertex fit was then performed .
an event was written to tape if all the fast tracks in the event were _ inconsistent _ with having come from a single primary vertex .
this filter passed 1/8 of all interaction triggers and had about @xmath17 efficiency for otherwise accepted charm decays .
the experiment recorded data from @xmath18 inelastic interactions and wrote @xmath19 events to tape using both positive and negative beams .
the sample was @xmath20 @xmath21-induced , with the balance split roughly equally between @xmath22 and protons .
the offline analysis selected single charm events with a topological identification procedure . only charged tracks with reconstructed momenta were used .
tracks which traversed the rich ( @xmath23 ) were identified as protons or kaons if those hypotheses were more likely than the pion hypothesis .
all other tracks were assumed to be pions .
the primary vertex was refit offline using all found tracks .
an event was rejected if all tracks were consistent with one primary vertex .
for those events which were inconsistent with a single primary vertex , secondary vertices were formed geometrically and then tested against a set of charge , rich identification , and mass conditions to identify candidates for the different single charm states .
candidate events were written to a charm data summary file .
subsequent analysis began by selecting particular single - charm species from that set of events .
in this study we began with the selex @xmath24 sample that has been used in lifetime and hadroproduction studies @xcite .
the sample - defining cuts are described in that reference .
no new cuts on the d mesons were introduced in this analysis .
the d meson momentum vector had to point back to the primary vertex with @xmath25 ( the double charm lifetime is known to be much shorter than the d meson lifetime , so the d meson pointback is not affected by having come from a secondary decay ) .
the d meson decay point must have a vertex separation significance of at least 10 @xmath26 from the primary . everywhere in these analyses
the vertex error used is the quadrature sum of the errors on the primary and secondary vertices .
the k was positively identified by the rich detector .
the pions were required to be rich - identified if they went into its acceptance .
the and mass distributions are shown in fig .
[ dmass ] .
there are 1450 decays and 2450 decays in these samples .
the events contribute to the signal channel .
the events can not come from the decay of a double charm baryon and will be used as a topological background control sample .
the yield asymmetry stems from the d - quark contribution of the @xmath9 beam component that gives a sizeable production asymmetry favoring leading @xmath27 production over @xmath28 production , as we have reported for other charm systems @xcite .
the track - based search code is identical to that used on the @xmath29 sample in the original investigation @xcite .
the premise is that a ccd state will make a secondary decay vertex between the primary production vertex in one of the thin foil targets and the observed d meson decay vertex , which must lie outside material .
we looked for intermediate vertices using all charge zero pairs of tracks from the set of reconstructed tracks not assigned to the d - meson candidate .
the additional positive track in this final state must be rich - identified as a proton if it traverses the rich .
the negative track in the new vertex is assigned the kaon mass .
( this track typically missed the rich acceptance by being too soft or too wide - angle , as confirmed by simulation . )
we made background studies by ( i ) assigning the negative track a pion mass , ( ii ) looking for proton plus positive track combinations with a d meson , and ( iii ) looking at proton @xmath30 combinations ( wrong - sign charm ) .
we require a good 3-prong vertex fit with a separation significance of at least 1.0 @xmath26 from the primary vertex , the same requirement used in ref .
the primary position was recalculated from the beam track and secondary tracks assigned to neither the d nor the p@xmath8 vertices .
results presented in this paper come from this analysis .
the signal search was based on a 10 window centered on the mass 3519 from ref .
the expected mass resolution for the decay is 4 .
our simulation correctly reproduces the observed widths of all our reported single charm mesons and baryons .
the 10 window should collect 80% of the events in a signal .
the results are insensitive to changing bin boundaries by up to half a bin .
the background , assumed to be flat , is evaluated outside a 20 window , to avoid putting the remaining 20% of any signal into the background .
the right - sign mass combinations in fig .
[ dpkmass](a ) show an excess of 5.4 events over a background of 1.6 events . the wrong - sign mass combinations ( @xmath31 quark in the decay ) for the @xmath32 final state
are also plotted in fig .
[ dpkmass ] ( b ) , scaled by 0.6 for the @xmath33ratio .
the wrong - sign background shows no evidence for a significant narrow structure near 3519 .
the average wrong - sign occupancy is 0.4 events / bin , exactly the background seen in the right - sign channel .
this confirms the combinatoric character of the background population in the right - sign signal .
we have investigated all possible permutations of particle assignments .
the only significant structure observed is in the channel , the place where a double charm baryon decay can occur . in order to assign a significance level to this peak
, we have combined statistical methods used in the original double charm paper @xcite and the event - mixing method used in ref .
we set out to test the hypothesis that the background events in fig .
[ dpkmass ] are random combinatoric tracks associated with real @xmath34 mesons . to mix events we took a @xmath34 meson in the peak region ( fig .
[ dmass ] ) and combined it with proton and @xmath35 tracks extracted from other events .
each @xmath34 was reused 25 times .
to compare to the combinatoric background in fig .
[ dpkmass ] , we scale the mixed - event background down for the multiple @xmath34 usage . the resulting background distribution predicts the observed distribution very well .
the mean number of background events below , in , and above the 4-bin signal peak is 5.74 @xmath4 0.26 , 1.38 @xmath4 0.13 , and 10.60 @xmath4 0.36 . this agrees well with the 8 , 1.6 , and 10 events that we observe in the corresponding regions . a mass plot with the combinatoric background level is shown in fig .
[ combo ] .
the background that we observe is completely consistent in shape and normalization with random combinatoric tracks associated with real @xmath34 mesons .
we take the combinatoric background to give a proper measure of the expected background under the 7-event signal region .
the single - bin significance of the signal , using the above background numbers , is : s = ( 7 - 1.38)/@xmath36 = 4.8 @xmath26 . the poisson probability of observing at least this much excess , including the gaussian uncertainty in the background , is 6.4 x @xmath37 .
both of these statistical significance calculations use methods identical to those in ref @xcite .
this indicates a robust signal atop a combinatoric background whose shape and normalization are very well understood .
in order to estimate the mass of the state in light of the sparse statistics in fig .
[ dpkmass ] , we fixed the width of the gaussian to 4 and fitted the data distribution around the signal peak .
the fit mass is @xmath4 .
this agrees beautifully with the measurement of 3519 @xmath4 2 from the original double charm baryon report .
we present these data as confirmation of the double charm state at 3520 @xmath7 in a new decay mode .
the weighted average mass is @xmath38 .
the mass distributions for the two channels are shown in fig .
[ dpkmass_both ] .
we have used the simulation to study the relative acceptance for the two decay channels and in order to quote a relative branching ratio .
the overall acceptance , including the single charm selection and the proton i d requirements in the mode , is very similar .
selex measures the relative branching ratio @xmath39()/@xmath39 ( ) @xmath40 .
the systematic error due to acceptances is well understood from single charm studies and is negligible compared to the statistical error . in ref .
@xcite we noted that all observed ccd events were produced by the baryon beams .
none came from pions . in this sample , 1 event out of the 7 in the peak region seen in fig .
[ dpkmass ] is a pion beam event , and 1 of the 19 sideband events comes from the pion beam sample .
this sample is consistent with the view that double charm baryons are produced dominantly by the baryon beams in selex . in another comparison
, we had noted that the decays had an exceptionally short reduced proper time distribution , indicating a decay lifetime 5 - 10 times shorter than the lifetime .
that feature is confirmed by the channel . as we noted in ref .
@xcite , our lifetime resolution is excellent but we can not exclude 0 lifetime ( strong decay ) for these events .
the width of this peak is completely consistent with simulation of a zero - width state , unlikely for a strong decay of a massive state .
also , we do not see an increase in the signal when we reduce the vertex significance cut l/@xmath26 below 1 .
if this were a strong decay , one would expect as many events with l/@xmath26 of -1 as + 1 , so the signal should grow significantly .
it does not . in ref .
@xcite we noted that the yield and acceptance implied that a large fraction of the decays seen in selex came from double charm decays .
that was a surprise .
for the case that is not true .
only a few percent of the selex @xmath28 events are associated with double charm .
in summary , selex reports an independent confirmation of the double charm baryon previously seen in the decay mode via the observation of its decay into the final state .
using only very loose cuts gives the statistically - significant signal shown in fig .
[ dpkmass ] .
a combinatoric background from event mixing describes the non - signal distribution well .
the gaussian signficance using this background estimate is @xmath2 .
this decay mode confirms that the lifetime is very short and that it is produced dominantly by baryon beams , as we reported in ref .
the authors are indebted to the staff of fermi national accelerator laboratory and for invaluable technical support from the staffs of collaborating institutions .
this project was supported in part by bundesministerium fr bildung , wissenschaft , forschung und technologie , consejo nacional de ciencia y tecnologa ( conacyt ) , conselho nacional de desenvolvimento cientfico e tecnolgico , fondo de apoyo a la investigacin ( uaslp ) , fundao de amparo pesquisa do estado de so paulo ( fapesp ) , the israel science foundation founded by the israel academy of sciences and humanities , istituto nazionale di fisica nucleare ( infn ) , the international science foundation ( isf ) , the national science foundation ( phy # 9602178 ) , nato ( grant cr6.941058 - 1360/94 ) , the russian academy of science , the russian ministry of science and technology , the secretara de educacin pblica ( mexico ) ( grant number 2003 - 24 - 001 - 026 ) , the turkish scientific and technological research board ( tbitak ) , and the u.s .
department of energy ( doe grant de - fg02 - 91er40664 and doe contract number de - ac02 - 76cho3000 ) . v.v .
kiselev , a.k .
likhoded , a.i .
onishchenko , phys .
* d60 * ( 1999 ) 014007 ; eur .
j. * c16 * ( 2000 ) 461 .
+ b. guberina , b. melic , h. stefancic , eur .
j. * c9 * ( 1999 ) 213 ; _ ibid _ , * c13 * ( 2000 ) 551 . selex collaboration , j.s .
russ _ et al .
_ , in _ proceedings of the 29th international conference on high energy physics
_ , 1998 , edited by a. astbury _
( world scientific , singapore , 1998 ) , vol .
ii , p. 1259 ; hep - ex/9812031 . | we observe a signal for the doubly charmed baryon in the decay mode to complement the previous reported decay in data from selex , the charm hadro - production experiment at fermilab . in this new decay mode
we observe an excess of @xmath0 events over a combinatoric background estimated by event mixing to be @xmath1 events .
the mixed background has gaussian statistics , giving a signal significance of @xmath2 .
the poisson probability that a background fluctuation can produce the apparent signal is less than @xmath3 .
the observed mass of this state is 3518 @xmath4 3 , consistent with the published result .
averaging the two results gives a mass of @xmath5 .
the observation of this new weak decay mode confirms the previous selex suggestion that this state is a double charm baryon .
the relative branching ratio for these two modes is 0.36 @xmath4 0.21 . the selex collaboration , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , doubly charmed baryon 14.20.lq , 14.40.lb , 13.30.eg |
previous works suggest that the most luminous extended emission line regions ( eelrs ) around type 1 quasars ( qsos ) at redshift @xmath30.5 exist preferentially associated with steep - spectrum radio - loud qsos with luminous [ oiii]@xmath05007 nuclear emission and low broad line region metallicities ( @xmath40.6@xmath5 ) .
these eelrs have low metallicities as well and seem especially prevalent in qsos showing signs of strong interaction .
the origin of these nebulae is controversial , whether cold accretion of intergalactic gas , tidal debris from a galactic merger or remnants of galactic superwinds ( for a review see fu & stockton 2009 ; stockton , fu & canalizo 2006 ) .
very little is known about the existence of eelrs associated with type 2 quasars ( e.g. humphrey et al .
2009 ; gandhi , fabian & crawford 1996 ) .
these objects are unique laboratories to investigate the existence and properties of quasar eelr .
the fortuitous occultation of the active galactic nucleus ( agn ) acts like a `` natural coronograph '' , allowing a detailed study of many properties of the surrounding medium , without the problems associated with the bright quasar point spread function . on the other hand , due to their much lower radio luminosity ,
distortions imprinted by the radio activity on the properties of the underlying nebulae are less important than in radio - loud objects . for these reasons
, we are undertaking a project based on imaging , long - slit and integral field data obtained with fors2/vlt , the integral field spectrograph pmas on the 3.5 m telescope at calar alto observatory ( humphrey et al .
2009 ) , and the tunable filter osiris on the gtc .
these are being used to investigate the existence of eelr associated with sloan digital sky survey ( sdss ) type 2 quasars at @xmath60.3 - 0.4 and characterize their morphological , ionization and kinematic properties . in this paper we present results on the type 2 quasar sdss j012341.47 + 004435.9 ( sdss j0123 + 00 hereafter ) at @xmath10.399 ( zakamska et al .
the object was detected by the nvss survey with a flux density of [email protected] mjy ( condon et al .
the radio emission is spatially unresolved . assuming a radio spectral index of @xmath8 ( where @xmath9 ) , the expected radio luminosity density at 5 ghz is 5@xmath1010@xmath11 erg s@xmath12 hz@xmath12 sr@xmath12 .
this value suggests that the quasar is radio - quiet " or at most radio - intermediate " ( miller , peacock & mead 1990 ) .
we assume @xmath13=0.7 , @xmath14=0.3 , h@xmath15=71 km s@xmath12 mpc@xmath12 . in this cosmology
, 1@xmath16 corresponds to 5.33 kpc at @xmath17=0.399 .
the vlt observations were carried out on the 8th and 9th sept 2008 using fors2 on ut1 .
the 691@xmath1855 + 69 filter was used to obtain the intermediate band image on the 1st night .
it has a spectral window 6635 - 7185 which covers both continuum and redshifted h@xmath19 and [ oiii]@xmath204959,5007 .
the pixel scale is 0.25 arcsec pix@xmath12 .
a diamond pattern mosaic of 4 images with individual exposures of 200 seconds was obtained , shifting each image by 10@xmath16 in ra or dec relative to the previous one .
this was done to minimize the impact of bad pixels and ghosts on the ccd .
the images were combined , with simultaneous cosmic ray removal after bias subtraction . the seeing size measured from the final image is [email protected]@xmath16 for the 1st night .
long slit spectroscopy was obtained at position angles ( pa ) pa60 ( first night ) and pa76 ( second night ) , aligned with extended diffuse emission or nearby potentially companion objects revealed by the images ( fig . 1 ) .
the 600ri+19 grism and the gg435 + 81 order sorting filter were used with a 1@xmath16 wide slit .
the useful spectral range was @xmath25030 - 8250 corresponding to rest frame 3595 - 5901 .
the object was moved along the slit between two different positions for a more accurate sky subtraction .
the data were reduced following standard procedures with starlink and iraf packages .
the spectra were debiased , flat fielded , wavelength ( @xmath0 ) calibrated , background subtracted , combined with simultaneous cosmic ray removal , flux calibrated and corrected for galactic extinction .
geometric distortion was found to be negligible .
the nights were photometric .
comparison of several spectrophotometric standard stars taken with a 5@xmath16 slit during the run gave a flux calibration accuracy of 5% over the entire spectral range . to perform the spectroscopic analysis ,
the line profiles were fitted with gaussian functions .
the fwhm values were corrected for instrumental broadening ( [email protected] as measured from the sky emission lines ) .
point sources did not fill the slit during the second night ( seeing [email protected]@xmath16 ) and this introduces uncertainties on the kinematic analysis .
for this reason , the pa60 spectrum was used to characterize the kinematic properties at the different spatial regions identified in the spectra .
gtc images were obtained with the osiris camera on 17th oct 2009 using the order sorting filter 694/44 .
the @xmath21 detector binning provided a spatial scale of 0.254 arcsec pix@xmath12 .
a continuum image and a continuum+[oiii]@xmath05007 image were produced .
each one consisted of three 200 second exposures obtained through the tunable filters with fwhm=16 and central @xmath0 ( @xmath22 ) at the optical axis of 7019 and 7069 respectively . due to the nature of the instrument @xmath22 changes with radial distance from the optical axis .
@xmath22 values were 7002.9 and 7052.7 for the on - band and off - band images respectively at the location of the target , corresponding to rest - frame [ oiii]@xmath05007 and the adjacent continuum .
the average seeing was [email protected]@xmath16 .
the observations were reduced and flux calibrated with the standard star g158 - 100 using iraf routines .
tunable filter profiles were generated at @xmath22 corresponding to the spatial location of g158 - 100 on the ccd and applied to its absolute flux spectrum to obtain the flux calibration factors .
the individual target exposures for each filter were sky subtracted , spatially aligned and combined .
the vlt intermediate band image is shown in fig . 1 ( top panel ) .
the bottom panel shows the gtc pure [ oiii]5007 ( green ) and continuum ( red ) composite image .
the original [ oiii]+continuum image was continuum subtracted , and therefore the green colour shows the pure gas emission .
the central elliptical galaxy presents distorted contours due to a detached `` blob '' which is more evident in the gtc images and emits both continuum and strong line emission .
low surface brightness diffuse features and knots are detected well outside the optical size of the galaxy .
a long filament extends roughly in the e - w direction , to a maximum projected distance from the nuclear region of @xmath2314@xmath16 ( 75 kpc ) and a total extension of 25@xmath16 ( 133 kpc ) ( throughout the text distances and sizes are projected values ) .
the filament is clearly detected in [ oiii ] extending towards and overlapping with galaxy `` g1 '' , which emits not only continuum but also line emission since it is clearly detected in the pure [ oiii ] gtc image .
therefore it is at similar @xmath17 as the quasar .
both objects seem to be physically connected by the filament mentioned above , which is reminiscent of a tidal bridge .
`` g2 '' , which is detected both in continuum and [ oiii ] , also lies at a similar @xmath17 and is located roughly along the direction of the filament .
other galaxies are found near the quasar in projection , but we can not confirm that any of them are at similar @xmath17 .
the vlt and gtc images show that sdss j0123 + 00 lies in a rich gaseous environment and suggest that it is a member of an interacting system .
the 2-dim spectra at pa60 and pa76 including the [ oiii]@xmath204959,5007 doublet are presented in fig . 2 .
the low surface brightness features revealed by the images ( fig .
1 ) along the slit direction emit a rich emission line spectrum .
continuum emission is not detected from these features .
the spectra reveal a total extension of the ionized nebula of 12.5@xmath16 ( 67 kpc ) and @xmath234@xmath16 ( 180 kpc ) along pa60 and pa76 respectively .
the maximum radial extent from the nucleus is @xmath2325@xmath16 ( 133 kpc ) measured along pa76 towards the east .
the filament revealed by the images is part of this structure .
the visual inspection of the 2-dim spectra reveal 3 distinct spatial regions where the gas is characterized by different kinematic and ionization properties , which are recognisable at both pa . we have named them the central region , the intermediate region and the low surface brightness nebula ( lsbn hereafter ) .
we have analysed them separately . since it is possible that within each region there is a spatial variation of the kinematic and ionization properties , we have defined within each region one or more apertures ( ap . hereafter ) as specified below . in order to characterize the ionization properties of the gas at each spatial location
, we plot the line ratios for the different apertures in the [ oiii]/h@xmath19 vs. [ oii]@xmath03727/[oiii]@xmath05007 diagnostic diagram ( fig .
3 ) , which involves the strongest emission lines detected in most apertures . only apertures for which at least two emission lines are detected are shown .
for comparison , the standard photoionization model sequence often applied to low redshift type 2 active galaxies is also shown ( see villar - martn et al .
2008 for a discussion of this model sequence applied to type 2 quasars ) .
the models assume solar metallicity , gas density @xmath24=100 @xmath25 and power - law index @xmath26=-1.5 .
the ionization parameter @xmath27 varies along the sequence .
a subsample of hii galaxies extracted from the catalogue by terlevich et al .
( 1991 ) is also shown .
* _ @xmath28 1@xmath16 .
the central region ( ap . 4 in fig . 2 and 3)_. this is the region in which the continuum emission is brightest
the continuum spatial profile is strongly peaked and is dominated by an unresolved central component ( [email protected]@xmath16 , pa60 ) , although diffuse continuum emission is also detected in excess of the psf wings up to a total extension of @xmath24@xmath16 or @xmath221 kpc .
the central region emits a rich emission line spectrum , typical of agn ( fig . 3 dark blue solid circle ) , including strong heii@xmath04686 .
the fwhm([oiii ] ) value measured from the 1-dim spectrum is 430@xmath730 km s@xmath12 .
+ we have fitted the continuum in ap . 4 stellar population models .
the nebular continuum was first removed .
its contribution , 6% of the total continuum flux at 4020 ( @xmath10 ) , was estimated from the h@xmath19 flux .
a reasonable range of parameters ( gas metallicities , temperature and density ) and no reddening ( as implied by the h@xmath19/h@[email protected] value ) were considered .
the fit ( fig . 4 ) to the nebular continuum subtracted
spectrum implies the presence of an old stellar population and an excess of blue light which can not be accounted for with only old stars .
if it is due to young stars , the best fit is obtained with an old stellar population of @xmath305 - 10 gyr of age that contributes with @xmath280% of the continuum at 4020 ( @xmath10 ) plus a young stellar population ( age@xmath31 10 myr ) which contributes @xmath220% . however , based on these data alone we can not distinguish if this blue excess is alternatively due to scattered nuclear radiation ( the polarization level measured by zakamska et al .
( 2006 ) was ( [email protected]% ) .
* _ 1 @xmath32 2.5@xmath16 .
the intermediate region ( ap . 3 and 5 fig . 2 and 3)_. it extends at both sides of the central region , up to a distance of 2.0@xmath16 to the e and @xmath22.6@xmath16 to the w respectively from the continuum centroid ( pa60 )
the location of ap . 3 and 5 in the diagnostic diagram ( fig .
3 , cyan solid square and triangle ) shows that they are predominantly photoionized by the agn .
this is also confirmed by the presence of strong heii emission .
+ an interesting feature within the intermediate region , which is clearly seen in the gtc images ( fig .
1 ) is a detached blob , which looks rather compact ( ap . 5 , fig . 2 ) .
fwhm values along pa60 ( [email protected]@xmath16 ) and pa76 ( [email protected]@xmath16 ) implying a physical size of a few kpc .
its centroid is located at @xmath21.3@xmath16 from the continuum centroid .
faint continuum emission is detected .
the lines are blueshifted relative to the rest of the ionized gas ( fig . 2 ) and
show very asymmetric line profiles which reveal complex kinematics . *
_ @xmath333@xmath16 . the low surface brightness nebula ( lsbn ) _ has a total extension of @xmath2180 kpc , as mentioned above ( fig .
2 ) , and overlaps towards the east with the filament or tidal bridge identified in the images ( fig . 1 ) .
the fact that the first and nvss fluxes are consistent with each other shows that no significant radio emission at 1.4 ghz is present on scales @xmath345@xmath16 or 27 kpc and therefore any jet present is probably much smaller than the nebula .
the spectrum integrated across ap .
6 to 9 along pa76 is shown in fig .
a study of the emission line kinematics of the lsbn along pa60 shows that they are quiescent : it emits very narrow , unresolved emission lines with fwhm@xmath35100 to @xmath35140 km s@xmath12 depending on spatial location , and the rest - frame velocity shifts relative to the nucleus are also small ( @xmath36200 km @xmath37 ) .
the lines are narrower than the ip measured from the sky lines at most spatial locations along pa76 , confirming the general narrowness of the lines also in this direction . + in order to investigate the spatial variation of the ionization properties of the lsbn , several apertures where considered along the slit . since the lsbn is knotty and inhomogeneous , we selected apertures that isolate individual features when possible ( fig . 2 ) .
we have plotted the location of the apertures along pa60 and pa76 in the diagnostic diagram ( fig .
it is important to note that the standard agn photoionization models , although successful at reproducing many spectral properties of type 2 active galaxies in general and type 2 quasars in particular , can not explain the large scatter in the diagnostic diagrams presented by this class of objects ( e.g. robinson et al .
1987 ) , with a substantial fraction overlapping with hii galaxies .
the locations of the lsbn apertures of sdss j0123 + 00 in the diagrams are consistent with the scatter presented by the general population of type 2 active galaxies . although a range of gas and/or agn continuum properties has usually been claimed as responsible for the scatter , villar - martn et al .
( 2008 ) proposed that an alternative possibility is that young stars contribute to the gas excitation , in addition to the agn .
we explore this possibility further for sdss j0123 + 00 .
3 shows that ap .
2 and 6 line ratios along pa60 , which are closer to the agn , are consistent with pure agn photoionization , while moving outwards ( ap . 1 , at @xmath236.6@xmath16 or 35 kpc ) , the line ratios are consistent with hii star forming galaxies .
a similar trend is found along pa76 : ap .
6 presents an intermediate spectrum , with line ratios between the standard agn models and hii galaxies ( heii is also detected ) .
all other apertures ( @xmath384.5@xmath16 or 24 kpc , most clearly ap
. 2 , 7 , 8 , 9 ) overlap with the hii galaxy zone .
+ therefore , although with the available information it is not possible to discard that the entire nebula is photoionized by the agn , we find that the spatial variation of the lsbn line ratios can be explained if the inner regions are photoionized by the agn , while moving outwards stellar photoionization becomes dominant .
line ratios based on [ oi]@xmath06300 , h@xmath26 , [ nii]]@xmath06583 , [ sii]@xmath206716,6732 would help to discriminate between the two scenarios . + in the stellar photoionization scenario , we can estimate the gas oxygen abundance in regions purely ionized by stars , using the r23 method ( pilyugin & thuan 2005 ) .
the spectrum integrated across the lsbn has h@xmath29/h@[email protected] , consistent within the errors with the theoretical case b recombination value , and thus we assume no line reddening . to make sure that agn photoionization is relatively negligible , we have used the spectra of those apertures along pa76 ( e.g. ap .
7,8,9 ) which are unambiguously placed among the hii galaxies and far from the standard agn models in fig . 3 , taking the errorbars into account .
+ the individual spectra , as well as the integrated spectra across all these apertures , are too noisy to set useful constraints on the [ oiii]@xmath04363 flux and @xmath39 .
this introduces uncertainties on the abundance determination , due to the bimodal shape of the 12+log(o / h ) vs. log(r23 ) function ( pilyugin & thuan 2005 ) .
the @xmath40 log values measured for ap . 7 , 8 an 9 along pa76 are [email protected] , [email protected] and [email protected] respectively and imply abundances well below solar , in the range 8% to 42% solar taking errors and uncertainties into account , and depending on the aperture .
similarly abundances are often found in compact hii galaxies ( e.g. guzmn et al .
1997 , terlevich et al .
1991 ) . for comparison
, we have used the sdss spectrum to constrain the abundance within the inner agn photoionized region using the method discussed in humphrey et al .
the resulting oxygen abundance must be @xmath3425% solar in order to reproduce the optical line ratios .
+ the narrowness of the spectral lines compared with the ip ( see 2 ) , implies that the spatial extent in the direction perpendicular to the slit is @xmath411@xmath16 or @xmath415.3 kpc . for the brightest knots isolated along the slit , [email protected]@xmath70.1@xmath16 corresponding to [email protected] kpc , taking the seeing into account .
sizes @xmath355 kpc are also measured for the uv rest frame continuum knots identified in the hst images of zakamska et al .
+ the h@xmath19 luminosity for ap .
7,8,9 is in the range ( 0.4 - 1.0)@xmath1010@xmath42 erg s@xmath12 .
if the gas is photoionized by stars , star forming rates ( sfr ) @xmath20.09 - 0.23 m@xmath43 yr@xmath12 are implied ( kennicutt 1998 ; given the large ew of the lines , stellar absorption of h@xmath19 is expected to be negligible ) . considering the small knot sizes , slit flux losses could be @xmath3530% .
this uncertainty does not affect our conclusions .
similar sfrs and sizes are found for some nearby compact hii galaxies ( telles , muoz - tun , tenorio - tagle 2001 , guzmn et al .
sdss0123 + 00 at @xmath10.399 is associated with a giant , luminous ionized nebula , which extends for @xmath34180 kpc .
this is one of the largest eelrs ever detected associated with an agn .
the eelr @xmath44}\sim$]3@xmath1010@xmath45 erg s@xmath12 is similar to that of very luminous eelrs associated with type 1 quasars .
therefore , this object demonstrates that very luminous and giant eelr can also be associated with low radio luminosity quasars , contrary to expectations based on type 1 quasar studies ( see 1 ) .
the low eelr abundances inferred ( @xmath410.5@xmath5 ) and the clear signs of interactions are features often found in type 1 quasars with luminous eelr .
sdss0123 + 00 lies in a gas rich environment and it seems to be a member of an interacting system .
the quasar appears to be physically connected by a tidal bridge to a companion galaxy at @xmath2100 kpc in projection .
the giant eelr encompasses the bridge . although we can not discard the possibility that the entire eelr is photoionized by the agn , the spatial variation of the line ratios can be explained if the gas in the inner regions ( @xmath4624 kpc ) is predominantly ionized by the agn while stellar photoionization becomes dominant as we move outwards .
the star formation might have been triggered during the interaction .
whatever ionizes the gas , we find that the properties ( sizes , line widths , line ratios and luminosities ) of the individual knots isolated within the lsbn are consistent with nearby hii galaxies .
based on the environmental properties ( the tidal bridge and the companion galaxy `` g1 '' ) , we propose that the origin of the eelr in this object is tidal debris from a galactic encounter .
the companion must contain a substantial amount of metal poor gas , to explain the subsolar eelr abundances .
the nuclear activity might have been triggered in such encounter , rather than in the final stages of a major galaxy merger ( as the nuclei coalesce ) .
probably the closest approach in the encounter occurred @xmath30100 myr ago , assuming a typical 300 km s@xmath12 separation speed , and a true distance between the two galaxies of 100 kpc .
such a scenario would be consistent with recent imaging results for powerful radio galaxies ( ramos almeida et et .
2010 , in prep . )
thanks to an anonymous referee for useful comments .
work funded with support from the spanish micinn ( grants aya2004 - 02703 , aya2007 - 64712 and aya2009 - 13036-c02 - 01 ) .
ams is supported by a uk stfc postdoctoral fellowship .
ep is grateful to ignacio gonzlez serrano for useful discussions and help with osiris data .
thanks to the gtc and vlt staff for their support during the observations . | we present long - slit spectroscopy , continuum and [ oiii]@xmath05007 imaging data obtained with the very large telescope and the gran telescopio canarias of the type 2 quasar sdss j0123 + 00 at @xmath10.399 .
the quasar lies in a complex , gas - rich environment .
it appears to be physically connected by a tidal bridge to another galaxy at a projected distance of @xmath2100 kpc , which suggests this is an interacting system .
ionized gas is detected to a distance of at least @xmath2133 kpc from the nucleus .
the nebula has a total extension of @xmath2180 kpc .
this is one of the largest ionized nebulae ever detected associated with an active galaxy .
based on the environmental properties , we propose that the origin of the nebula is tidal debris from a galactic encounter , which could as well be the triggering mechanism of the nuclear activity .
sdss j0123 + 00 demonstrates that giant , luminous ionized nebulae can exist associated with type 2 quasars of low radio luminosities , contrary to expectations based on type 1 quasar studies .
( galaxies : ) quasars : emission lines ; ( galaxies : ) quasars individual : sdss j012341.47 + 004435.9 ; galaxies : ism ; [ firstpage ] |
complex networks are often employed as models for large - scale systems like connectivity inside the brain , linking structure of the internet or trust relations in social networks . even in cosmology
, causality can be modeled based on a network as demonstrated in @xcite .
such networks are of hyperbolic structure , in which older nodes are favorably connected compared to younger ones . in many settings
the interconnections of a network contain uncertainties or represent the ( possibly unknown ) parametrized quantities of interest .
parametrized models with high - dimensional state and parameter spaces are often infeasible to evaluate many times for different locations of the parameter space .
this is due to two effects .
first , the high - dimensional state space makes each integration of the dynamic system computationally costly .
second , the high - dimensional parameter space may make many simulations necessary .
in this situation model reduction will accelerate these otherwise costly experiments .
particularly , the combined reduction of state and parameter space will be illustrated .
this setting for model reduction was inspired by @xcite .
the gramian - based ( state ) reduction approach originates in ( approximate ) balanced truncation comprehensively described in @xcite .
an alternative computational method for these gramians , based on proper orthogonal decomposition , was introduced in @xcite under the name * empirical gramians*. for the parameter identification and combined state and parameter reduction , the empirical joint gramian from @xcite is utilized . in the next section the construction of a hyperbolic network
is described . in section [ sr ]
the state reduction procedure , then in section [ pr ] the parameter identification and combined state and parameter reduction is explained . for an efficient assembly of the required gramians , the empirical cross gramian is presented in section [ eg ] . in section
[ uq ] the usage of empirical gramians in the context of uncertainty quantification is outlined .
finally , in section [ nr ] a sample network is reduced .
generating a hyperbolic network is a dynamic process with a discrete time space .
the following description is taken from @xcite and ( * ? ? ?
* supplementary notes ii .
c ) . at each time step
@xmath0 a new node is born by drawing from a uniform random distribution on the circle @xmath1 yielding a new node at @xmath2 $ ] and a radius @xmath3 with network degree @xmath4 .
the new node @xmath5 connects to all existing nodes @xmath6 that satisfy : @xmath7 this leads to a `` space - time '' representation in which the network is created in the shape of a hyperboloid visualized in figure [ hy ] . in this work
a maximum number of @xmath8 nodes will be set .
such a hyperbolic network can be modeled as a matrix @xmath9 , by treating a matrix element @xmath10 as connection from node @xmath11 to @xmath12 , which leads to a dynamic system setting of a basic linear autonomous system : @xmath13 with state @xmath14 and a system matrix @xmath9 embodying the network structure .
usually some external input or control is applied to the system and the quantities of interest are some subset or linear combination of the systems states .
this leads to a linear control system @xmath15 with input @xmath16 , input matrix @xmath17 , outputs @xmath18 and output matrix @xmath19 .
the matrices @xmath20 can for example be used to excite only a certain nodes via @xmath21 and observe the dynamics in others via @xmath22 . in this setting
it is assumed , the connections , being the components of @xmath23 , between the networks nodes , are parametrized in each component by @xmath24 : @xmath25 as described above , in each time step a new node is born and connects to existing nodes . hence , the components of @xmath23 , which are components of @xmath26 , too , change over time and yield a parametrized * linear time - varying control system * : @xmath27 now , the hyperbolic network can be treated with the model reduction methods of control theory .
a well known method for model order reduction of the state space in a control system setting is * balanced truncation*. in this approach a systems controllability and observability is balanced and the least controllable and observable states are truncated .
a systems controllability is encoded in a gramian matrix @xmath28 which is a solution of the lyapunov equation @xmath29 .
a systems observability is also encoded in a gramian matrix @xmath30 which is a solution of the lyapunov equation @xmath31 .
then , by a balancing transformation , computed from the controllability and observability gramian , the control system is transformed in a manner such that the states are ordered from the most to the least important , for the systems dynamics .
this ordering is based on the hankel singular values @xmath32 .
after the sorting , the least controllable and observable can be truncated .
the * cross gramian * encodes controllability and observability into one gramian matrix .
a solution of the sylvester equation @xmath33 yields the cross gramian matrix @xmath34 as a solution , if the system is square . in case the system is symmetric , meaning the systems gain @xmath35 is symmetric , then the absolute value of the cross gramians eigenvalues equal the hankel singular values : @xmath36 given an asymptotically stable system , the cross gramian can also be computed as the time integral over the product of input - to - state and state - to - output map : @xmath37 which will be the basis for computing the empirical gramian variant . the state reduction is based on the singular values of the cross gramian @xmath34 . a singular value decomposition of @xmath38 , provides a projection of the states @xmath39 in which the states are sorted by their importance . without loss of generality , the singular values , composing the diagonal matrix @xmath40 , are assumed to be sorted in descending order .
based on this projection , the matrices @xmath41 and the initial value @xmath42 can be partitioned and reduced , @xmath43 this direct truncation approximates closely the balanced truncation of controllability and observability gramians , but does not require an additional balancing transformation . in case the system is not square or not symmetric , following the approach from @xcite , the system @xmath44 can be embedded into a symmetric system @xmath45 . since for each square matrix @xmath9 there exists a symmetrizer @xmath46 such that @xmath47 and thus the embedding system is given by : @xmath48 if the system matrix @xmath23 is symmetric , and thus @xmath49 , the embedding of the system simplifies to : @xmath50 even though the number of inputs and outputs is increased the number of states remains the same as in the original system .
the concept of controllability and especially observability extends to parametrized systems by treating the parameters as additional states .
these parameter states are constant over time and are assigned the parameters value as initial states : @xmath51 this augmented system , used in @xcite , can now be subject to a similar method to the direct truncation of the cross gramian for state reduction . the cross gramian of the augmented system yields the * joint gramian * introduced in @xcite : @xmath52 with its upper left block ( @xmath34 ) being the usual cross gramian of the system . the identifiability information of the parameters is encoded in @xmath53 . the parameter related information is then extracted by the schur - complement of the symmetric part of the joint gramian , resulting in the * cross - identifiability gramian * @xmath54 : @xmath55 a singular value decomposition of @xmath54 , provides a projection of the parameters that are sorted by their importance : @xmath56 based on this projection the parameters can be partitioned and reduced , @xmath57 as described in @xcite , by employing a truncation of states based on the singular values of @xmath34 and parameters based on the singular values of @xmath54 enables the combined reduction .
empirical gramians were introduced in @xcite and are solely based on simulations of the underlying control system .
these simulations use perturbations in input @xmath58 and initial states @xmath42 which are averaged .
the required perturbations are organized into sets allowing a systematic perturbation of input @xmath59 and initial states @xmath60 : @xmath61 now the empirical cross gramian can be defined as follows ( taken from @xcite ) : for sets @xmath62 , @xmath63 , @xmath64 , @xmath65 , @xmath66 , @xmath67 , input @xmath68 during steady state @xmath69 with output @xmath70 , the * empirical cross gramian * @xmath71 relating the states @xmath72 of input @xmath73 to output @xmath74 of @xmath75 , is given by : @xmath76 the joint gramian @xcite encapsulates the cross gramian , hence the * empirical joint gramian * is computed in the same manner as the empirical cross gramian , yet of the augmented system .
as shown in @xcite , the empirical gramians extend to time - varying systems , and thus can be applied in this setting for the hyperbolic networks .
the connections between the network nodes , which are modeled by the components of the system matrix @xmath23 , might contain uncertainties .
due to the computation of empirical gramians based on simulations , potential uncertainties in initial state and external input can be incorporated by enlarging the corresponding set of perturbations respectively .
hence , for an augmented system uncertainties in the parameters can also be included .
this allows robust model reduction .
additionally , the parameter reducing projection can also be used to reduce , for example in a gaussian setting , mean and covariance of a parameter distribution .
to demonstrate the capabilities of this approach a synthetic hyperbolic network is utilized .
as described in section [ hn ] , the time varying system is growing with each time step .
this network with a maximum of @xmath77 nodes , thus a state dimension of @xmath78 , and @xmath79 inputs and outputs is selected .
furthermore , it is assumed that each connection is reciprocal , hence @xmath80 and @xmath81 .
all possible connections of all nodes are treated as ( time - varying ) parameters in this setting , @xmath27 yet input matrix @xmath82 and output matrix @xmath83 are random and notably @xmath84 , which requires an embedding into a symmetric system ( see section [ sr ] ) . first , in an offline phase , that has to be performed only once , a reduced order model is created .
the reduction procedure uses the empirical joint gramian of section [ eg ] that computes the cross gramian of the embedded augmented system : @xmath85 then , a reduction of states , based on the singular values of @xmath34 , and of parameters , based on the singular values of @xmath54 , is performed .
second , in the online phase , the reduced model can be evaluated .
the computations are performed using the empirical gramian framework - * emgr * described in @xcite .
source code for the following experiments can be found at .
since the network evolves during its evaluation the reduction has to be performed for unknown connectivity .
a distribution of the singular values of the empirical cross gramian @xmath34 and the empirical cross - identifiability gramian @xmath54 is given in figure [ sv0 ] and figure [ sv1 ] .
the singular values of these two empirical gramians describe the energy contained in the state and parameter respectively .
a reduction of the parameter space from dimension @xmath86 to @xmath87 is suggested by the singular values of the cross - identifiability gramian . for the state space a reduction from dimension @xmath77 to @xmath88 is performed , based on the singular values of the cross gramian .
the reduced model can be evaluated and compared to the full order model .
figures [ hn0 ] and [ hn1 ] show the impulse response of the full order and reduced order network , while the relative error between them is shown in figure [ hn2 ] . between the two time series of the full and reduced order model impulse response , the relative @xmath89-error is @xmath90 .
the sharp drop in the singular values of the cross - identifiability gramian determines the reduced order model dimension ; a truncation of more parameter space dimensions will introduce a significant higher error .
the descent of singular values of the cross gramian also allows a graduated increase of error when truncating states . to scan various location of the parameter space
, only the low - dimensional reduced parameter space has to be scanned . comparing the reduced order model with the full order model ,
the combined reduction decreased the integration time by 21% and memory requirements by 70% .
.original time , offline time , online time and relative l2-error ; averaged over 100 simulations .
[ cols= " < , < , < , < " , ]
the numerical experiments suggest that the empirical joint gramian @xcite , which is based on the empirical cross gramian , can be applied to reduce this type of control system with hyperbolic network structure . as shown ,
a linear time - varying control system in which also the parameter values vary over time , can be handled by the empirical gramians .
this concurrent reduction of state and parameter spaces enables , for example , scenarios in which the reduced model can be used to scan the parameter space . using the ( empirical ) cross gramian of the embedded system is efficient , since the number of inputs and outputs is small compared to the number of states and no symmetrizer needs to be computed and inverted , @xmath23 has been chosen to be symmetric . yet , for networks with a non - symmetric system matrix @xmath23 , a possibly costly computation of the symmetrizer and its inverse is required .
thus , a generalization of the cross gramian to non - symmetric systems should be explored . | * we recently introduced the joint gramian for combined state and parameter reduction [ c. himpe and m. ohlberger .
cross - gramian based combined state and parameter reduction for large - scale control systems .
arxiv:1302.0634 , 2013 ] , which is applied in this work to reduce a parametrized linear time - varying control system modeling a hyperbolic network .
the reduction encompasses the dimension of nodes and parameters of the underlying control system .
networks with a hyperbolic structure have many applications as models for large - scale systems .
a prominent example is the brain , for which a network structure of the various regions is often assumed to model propagation of information .
networks with many nodes , and parametrized , uncertain or even unknown connectivity require many and individually computationally costly simulations . the presented model order reduction enables vast simulations of surrogate networks exhibiting almost the same dynamics with a small error compared to full order model . * +
* keywords : * hyperbolic network , model reduction , combined reduction , cross gramian , joint gramian , empirical gramian |
spherically symmetric geometries have a long and useful history as a physically interesting and technically vastly simplified arena for gravitational physics . in vacuum , einstein s theory with spherical symmetry has no local degrees of freedom , and the reduced phase space in the hamiltonian formulation is finite dimensional . including an idealized , infinitesimally thin matter shell brings in an additional finite number of degrees of freedom . including a continuous matter distribution generically yields a ( 1 + 1)-dimensional field theory , with the exception of fields whose gauge symmetries exclude spherically symmetric local degrees of freedom .
a familiar example of a field with such a gauge symmetry is the electromagnetic field . in this paper
we consider spherically symmetric einstein gravity coupled to an infinitesimally thin null - dust shell . from the spacetime point of view , the solutions to this system are well known ( see , for example , refs.@xcite ) , and they can be easily obtained from a junction condition formalism that is general enough to encompass null shells ( see ref .
@xcite and the references therein ) .
our purpose is to explore the hamiltonian structure of this system , treating both the geometry and the shell as dynamical . among the extensive previous work on hamiltonian approaches to spherically symmetric geometries ( for a selection in a variety of contexts , see refs . )
, we follow most closely the canonical transformation techniques of kucha @xcite .
our main results can be concisely described as generalizing the spherically symmetric vacuum hamiltonian analysis of ref.@xcite to accommodate a null - dust shell .
finding a suitable action principle requires care .
the shell stress - energy tensor is a delta - distribution with support on the shell history , which is a hypersurface of codimension one .
einstein s equations for the system therefore admit a consistent distributional interpretation @xcite , and the content of these equations is captured by the junction condition formalism of barrabes and israel @xcite .
we recover these equations from a variational principle .
we take the shell action to be that of a spherically symmetric thin cloud of radially - moving massless relativistic point particles , and we vary the total action independently with respect to the gravitational variables and the shell variables .
we shall see that this variational principle can be made distributionally consistent and that the variational equations do reproduce the correct dynamics . achieving this requires , however , a judicious choice of the regularity properties of the metric .
we begin , in section [ sec : metric ] , by setting up the hamiltonian formulation of the system in the arnowitt - deser - misner ( adm ) gravitational variables .
the spacetime is taken to have kruskal - like topology , with two asymptotically flat infinities , and the spatial hypersurfaces are taken to be asymptotic to hypersurfaces of constant killing time at each spacelike infinity .
the killing time evolution of the hypersurfaces is prescribed independently at each infinity .
we specify the regularity properties of the gravitational variables , and demonstrate that the variational principle is consistent and leads to the correct equations of motion . in section [ sec : transformation ]
we perform a canonical transformation to a new chart in which the constraints become exceedingly simple .
two of our new variables are hawking s quasilocal mass @xmath4 and the two - sphere curvature radius @xmath5 , just as in the vacuum analysis of ref.@xcite .
however , to maintain a consistent distributional interpretation of the variables in the new chart , we are led to relate the momentum conjugate to @xmath4 to the time whose constant value hypersurface coincides with the classical shell history , and not to the killing time as in ref .
the momentum conjugate to @xmath5 needs to be modified accordingly .
remarkably , the canonical transformation can then be chosen to leave the shell canonical pair invariant .
the transformation is mildly singular for geometries in which the shell straddles a horizon , but it can be extended to this special case in a suitable limiting sense . in section [ sec : reduction ]
we eliminate the constraints by hamiltonian reduction .
the reduced phase space @xmath0 turns out to have dimension four .
as the vacuum theory under our boundary conditions has a two - dimensional reduced phase space @xcite , and as a test shell in a fixed spherically symmetric background has a two - dimensional phase space , this is exactly what one would have anticipated .
we first obtain canonical coordinates @xmath6 in which the configuration variables @xmath7 are the schwarzschild masses on the two sides of the shell .
the momenta @xmath8 can be interpreted as the time differences between the shell and the infinities , after introducing an appropriate correspondence between our spatial hypersurfaces and hypersurfaces that are asymptotically null .
the configuration variables @xmath7 are constants of motion , while the shell motion is indirectly encoded in the dynamics of @xmath8 .
these coordinates become singular for horizon - straddling shells , but a global chart covering also this special case can be obtained by introducing suitable new momenta .
we find that @xmath0 consists of two disconnected copies of @xmath1 , each associated with one direction of the shell motion .
the right - moving and left - moving test shell limits , in which the shell stress - energy tensor vanishes , can be attached to the respective components of @xmath0 as smooth boundaries with topology @xmath2 . in section [ sec :
gauges ] we introduce on @xmath0 a local canonical chart in which the shell motion becomes more transparent . assuming that the shell does not a straddle a horizon ,
the shell history divides the spacetime into the interior , " which contains a killing horizon bifurcation two - sphere , and the exterior , " which does not .
we choose the configuration variables in the new chart to be the curvature radius of the shell two - sphere and the interior mass , in an arbitrarily specifiable stationary exterior coordinate system .
one can argue that this yields a hamiltonian description of interest for an observer who scrutinizes the shell motion from the exterior asymptotic region , especially if the observer s ignorance of the interior asymptotic region is incorporated by setting the interior contribution to the hamiltonian to zero .
we give three examples of stationary exterior coordinate systems in which the hamiltonian can be found in closed form .
also , choosing the spatially flat exterior gauge @xcite , and performing a partial reduction by setting the interior mass equal to a prescribed constant , we reproduce the spatially flat shell hamiltonian previously derived in refs.@xcite by different methods .
in section [ sec : para - times ] we free the evolution of the spatial hypersurfaces at the spacelike infinities by introducing parametrization clocks .
we find on the unreduced phase space a canonical chart in which the physical degrees of freedom and pure gauge degrees of freedom are completely decoupled , in full analogy with the vacuum analysis of ref.@xcite .
the pure gauge chart can be chosen so that the configuration variables are the curvature radius of the two - sphere and the time , with the latter one appropriately interpreted across the horizons . in section [ sec : flat - int ] we replace the kruskal spatial topology @xmath9 by the spatial topology @xmath2 .
the spacetime has then just one asymptotic region , and when the equations of motion hold , the spacetime interior to the shell is flat .
as in the kruskal case , we take the asymptotic region to be asymptotically flat , and we prescribe the evolution of the spatial hypersurfaces at the spacelike infinity
. we then carry out the canonical transformation and hamiltonian reduction .
expectedly , the reduced phase space turns out to consist of two disconnected copies of @xmath10 , with only the counterpart of the pair @xmath11 of the kruskal theory surviving .
we conclude in section [ sec : discussion ] with a summary and a brief discussion , including remarks on the potential utility of the results in view of quantization . some of the technical detail of the adm dynamical analysis is postponed to the appendices .
we work in planck units , @xmath12 .
lowercase latin tensor indices @xmath13 are abstract spacetime indices .
dirac s delta - function is denoted by @xmath14 , while @xmath15 denotes a variation .
the curvature coordinates @xmath16 for the schwarzschild metric are coordinates in which the metric reads @xmath17 where @xmath18 is the metric on the unit two - sphere and @xmath19 is the schwarzschild mass .
@xmath20 and @xmath21 are called respectively the killing time and the curvature radius .
in this section we set up the hamiltonian formulation for spherically symmetric einstein gravity coupled to a null - dust shell .
we pay special attention to the regularity of the gravitational variables and the global boundary conditions .
our spacetime geometry is given by the general spherically symmetric arnowitt - deser - misner ( adm ) metric @xmath22 where @xmath18 is the metric on the unit two - sphere , and @xmath23 , @xmath24 , @xmath25 , and @xmath21 are functions of the coordinates @xmath26 and @xmath27 only .
partial derivatives with respect to @xmath26 and @xmath27 are denoted respectively by overdot and prime , @xmath28 and @xmath29 .
we take the spacetime metric to be nondegenerate , and @xmath23 , @xmath25 , and @xmath21 to be positive .
the matter consists of an infinitesimally thin shell of dust with a fixed total rest mass @xmath30 , which we initially take to be positive . denoting the shell history by @xmath31 ,
the lagrangian matter action is @xmath32 where a hat is used to denote the value of a variable at the shell .
the shell can be envisaged as a thin spherically symmetric cloud of radially - moving massive relativistic point particles .
the lagrangian gravitational action is obtained by performing integration over the angles in the einstein - hilbert action , @xmath33 .
discarding a boundary term , the result is @xcite @xmath34 ( { \dot r } - r ' n^r ) + \case{1}{2 } \lambda { ( { \dot r } - r ' n^r ) } ^2 \right\ } \nonumber \\ & & + n \left ( \lambda^{-2 } r r ' \lambda ' - \lambda^{-1 } r r '' - \case{1}{2 } \lambda^{-1 } { r'}^{2 } + \case{1}{2 } \lambda
\right ) \bigg ] \ \ .
\label{s - lag}\end{aligned}\ ] ] the lagrangian action of the coupled system is @xmath35 we shall consider the regularity properties of the variables , the boundary conditions , and boundary terms after passing to the hamiltonian formulation .
the momenta conjugate to the configuration variables @xmath36 , @xmath25 , and @xmath21 are @xmath37 \ \ .
\label{pr}\end{aligned}\ ] ] a legendre transformation gives the hamiltonian bulk action @xcite @xmath38 \
\ , \label{s - ham}\ ] ] where the super - hamiltonian constraint @xmath39 and the radial supermomentum constraint @xmath40 contain both gravitational and matter contributions . in the limit @xmath41
, these constraints take the form [ metric - constraints ] @xmath42 where @xmath43 . from now on
, we shall work exclusively in this zero rest mass limit , with the bulk action ( [ s - ham ] ) and the constraints ( [ metric - constraints ] ) .
as will be verified below , the shell then consists of null dust .
the hamiltonian constraint ( [ superham ] ) is not differentiable in @xmath44 at @xmath45 .
as we shall verify below , an initial data set with nonzero @xmath44 can not evolve into a set with @xmath45 @xcite .
we assume from now on that @xmath44 is nonzero : this breaks the phase space into the two disconnected sectors @xmath46 .
the limits @xmath47 within each sector will be addressed in subsection [ subsec : reduction - no - shell ] . in the presence of a smooth matter distribution
, one can assume the spacetime metric to be smooth ( @xmath48 ) .
in the idealized case of an infinitesimally thin shell , the metric can be chosen continuous but not differentiable at the shell @xcite .
the issue for us is to find smoothness assumptions that give a consistent variational principle .
we wish to make both the action ( [ s - ham ] ) and its local variations well defined and such that the resulting variational equations are equivalent to einstein s equations with a null - dust shell .
we follow the massive dust shell treatment of ref . @xcite .
in contrast to the case of a massive dust shell , we shall find that the smoothness conditions introduced in ref .
@xcite make our null - dust variational principle fully consistent . as in ref .
@xcite , we assume that the gravitational variables are smooth functions of @xmath27 , with the exception that @xmath49 , @xmath50 , @xmath51 , @xmath52 , @xmath53 , and @xmath54 may have finite discontinuities at isolated values of @xmath27 , and that the coordinate loci of the discontinuities may be smooth functions of @xmath26 .
all the terms under the @xmath27-integral in the action ( [ s - ham ] ) are well defined in the distributional sense .
the most singular contributions are the explicit matter delta - contributions in the constraints , and the implicit delta - functions in @xmath55 and @xmath56 .
all these delta - functions are multiplied by continuous functions of @xmath27 .
the remaining terms are at worst discontinuous in @xmath27 .
the action is therefore well defined .
local independent variations of the action with respect to the gravitational and matter variables give the constraint equations [ metric - constr - eoms ] @xmath57 and the dynamical equations [ metric - dyn - eoms ] @xmath58 - { n ' r r ' \over \lambda^2 } + n^r p_\lambda ' \ \ , \label{plambda - eom }
\\ { \dot p}_r & = & n \left [ { \lambda p_\lambda^2 \over r^3 } - { p_\lambda p_r \over r^2 } - \left({r ' \over \lambda}\right ) ' \right ] - \left ( { n ' r \over \lambda}\right ) ' + { \left ( n^r p_r \right ) } ' \ \ , \label{pr - eom } \\ { \dot { \zr } } & = & { \eta { \hat n } \over { \hat \lambda } } - { \widehat{n^r } } \ \ , \label{rhat - eom } \\ { \dot { \zp } } & = & \zp { \left[\left ( n^r- { \eta n \over \lambda}\right)'\right]}^{\outwidehat } \ \ .
\label{rhatmom - eom}\end{aligned}\ ] ] as will be discussed in detail in appendix [ app : dist - eoms - gen ] , our smoothness conditions imply that equations ( [ metric - constr - eoms ] ) and all save the last one of equations ( [ metric - dyn - eoms ] ) have an unambiguous distributional interpretation .
equation ( [ rhatmom - eom ] ) , on the other hand , is ambiguous : the right - hand side is a combination of spatial derivatives evaluated at the shell , but these derivatives may be discontinuous .
we need to examine the dynamical content of the well - defined equations , and the possibilities of interpreting equation ( [ rhatmom - eom ] ) .
a first observation from equation ( [ rhat - eom ] ) is that the shell history is tangent to the null vector @xmath59 whose components are [ ell - def ] @xmath60 for @xmath61 ( @xmath62 ) , @xmath59 is the future null vector that points towards relatively larger ( smaller ) values of @xmath27 . from the definition of the shell stress - energy tensor , @xmath63 , we find @xmath64 the shell is therefore null , with positive surface energy but vanishing surface pressure @xcite .
this confirms that the shell consists of null dust .
all solutions to the spherically symmetric einstein equations with a null - dust shell can be found from a sufficiently general junction condition formalism @xcite . on each side of the shell ,
the spacetime is locally part of the extended schwarzschild geometry .
if the global structure of the spacetime is kruskal - like , with two asymptotically flat infinities , there are only two qualitatively different cases .
first , if the shell is not static , the junction is completely determined by continuity of the two - sphere radius at the shell .
the motion is clearly geodesic in each of the two geometries , and the radius of the two - sphere serves as an affine parameter in either geometry .
the spacetime is either that shown in figure [ fig : nonhorizon ] , or its time and/or space inverse .
second , if the shell is static , the junction is along a common horizon , and the masses must agree .
the soldering is affine , meaning that the affine parameters along the horizon with respect to the two geometries are affinely related ; however , as the stress - energy tensor of the shell is by assumption nonvanishing , the bifurcation two - spheres on the two sides do not coincide .
the spacetime is either that shown in figure [ fig : horizon ] or its time inverse .
now , away from the shell , equations ( [ metric - constr - eoms ] ) and ( [ metric - dyn - eoms ] ) are well known to be equivalent to einstein s equations .
we shall investigate equations ( [ metric - constr - eoms ] ) and ( [ metric - dyn - eoms ] ) at the shell in detail in the appendices .
the result is that , when combined with the fact that the geometry is locally schwarzschild on each side of the shell , the well - defined equations , ( [ metric - constr - eoms ] ) and ( [ lambda - eom])([rhat - eom ] ) , are equivalent to the correct null - dust junction conditions at the shell . they further imply that the right - hand side of ( [ rhatmom - eom ] ) is unambiguous , and that ( [ rhatmom - eom ] ) is satisfied as an identity .
our variational principle is therefore consistent , and it correctly reproduces the motion of a null - dust shell .
a check on the consistency of our formalism is that the poisson brackets of our constraints can be shown to obey the radial hypersurface deformation algebra @xcite , as in the absence of the shell , and as with a massive dust shell @xcite .
we therefore have a hamiltonian system with first class constraints @xcite .
what remains are the global boundary conditions .
we take the coordinate @xmath27 to have the range @xmath65 , and as @xmath66 we assume the falloff @xcite [ metricfall ] @xmath67 where @xmath68 and @xmath69 are functions of @xmath26 , and @xmath70 is a parameter that can be chosen freely in the range @xmath71 . here
, @xmath72 stands for a term that falls off as @xmath66 as its argument , and whose derivatives with respect to @xmath27 and @xmath26 fall off accordingly .
these conditions imply that the asymptotic regions associated with @xmath66 are asymptotically flat , with the constant @xmath26 hypersurfaces asymptotic to hypersurfaces of constant minkowski time .
@xmath69 are the rates at which the asymptotic minkowski times evolve with respect to the coordinate time @xmath26 . when the equations of motion hold , @xmath68 are time - independent and equal to the schwarzschild masses .
in the variational principle , we take @xmath69 to be prescribed functions of @xmath26 , but leave @xmath68 free .
the appropriate total action then reads @xcite [ total - metric - action ] @xmath73 where the boundary action is @xmath74 the global structure of the spacetime is kruskal - like , with two asymptotically flat asymptotic regions .
the classical solutions under these boundary conditions are precisely those described above and shown in figures [ fig : nonhorizon ] and [ fig : horizon ] .
in this section we find a new canonical chart in which the constraints become exceedingly simple .
away from the shell , our treatment closely follows that given by kucha in the vacuum case @xcite .
the new elements arise mainly from patching the two vacuum regions together at the shell .
our canonical transformation turns out to be mildly singular when the masses on the two sides of the shell agree and the shell straddles a common horizon .
we first perform the transformation , in subsection [ subsec : trans - nonhorizon ] , assuming that this special case has been excluded . we then argue , in subsection [ subsec : trans - horizon ] , that the transformation can be extended to the special case in a suitable limiting sense . in the vacuum theory ,
kucha @xcite found a transformation from the canonical chart @xmath75 to the new canonical chart @xmath76 defined by [ trans ] @xmath77 \ \ , \label{trans - psfr}\end{aligned}\ ] ] where @xmath78 when the equations of motion hold , @xmath19 is independent of both @xmath27 and @xmath26 , and its value is just the schwarzschild mass .
similarly , when the equations of motion hold , we have @xmath79 , where @xmath20 is the killing time .
the vacuum constraints can be written as a linear combination of @xmath80 and @xmath81 , and the dynamical content of the theory becomes transparent . in the presence of our null shell ,
the variables @xmath82 become singular at the shell . to see this , consider a classical solution in which the shell history does not lie on a horizon . as @xmath19 is discontinuous at the shell
, @xmath83 contains at the shell a delta - function in @xmath27 . as @xmath84 is discontinuous at the shell ,
the product @xmath85 is ambiguous .
one therefore does not expect the chart @xmath86 to be viable in the presence of the shell . to overcome this difficulty
, we keep @xmath19 and @xmath87 but replace the momenta by ones that are smoother across the shell .
we define first @xcite @xmath88 and @xmath89 note that @xmath90 .
when the equations of motion hold , @xmath91 vanishes on the leftgoing branch(es ) of the horizon and @xmath92 vanishes on the rightgoing branch(es ) of the horizon .
it follows that @xmath93 is nonvanishing on the horizon that the shell crosses .
now let [ pi - def ] @xmath94 when the equations of motion hold , equation ( [ pim - def ] ) shows that @xmath95 where @xmath96 is the tortoise coordinate @xcite . for @xmath97 ( @xmath98 )
, @xmath99 is the retarded ( advanced ) time coordinate .
while @xmath84 was associated with the killing time @xcite , our prospective new momentum @xmath100 is therefore associated with the retarded or advanced time .
away from the shell , a calculation of the poisson brackets shows that the set @xmath101 is a candidate for a new gravitational canonical chart .
we need to find shell variables that complete this set into a full canonical chart . for the rest of this subsection
, we assume that the shell history does not lie on a horizon .
the special case where the shell straddles a horizon will be discussed in subsection [ subsec : trans - horizon ] . as a preliminary , suppose that the constraints ( [ metric - constr - eoms ] ) hold , and consider the regularity of the variables .
away from the shell , the constraints ( [ metric - constr - eoms ] ) imply that @xmath52 and @xmath53 are continuous , and @xmath102 are thus both continuous . in the notation of appendix [ app : dist - eoms - gen ] , the distributional content ( [ delta - constraints ] ) of the constraints at the shell can be written as [ delta - constraints2 ]
@xmath103 from ( [ delta - constraints2-f ] ) we see that @xmath93 is continuous at the shell . equation ( [ pim - def ] ) then implies that @xmath100 is continuous , with the exception that it diverges on the horizon that is parallel to the shell history . the first equality sign in ( [ pisfr - def ] ) , and the observation that the vacuum constraints are linear combinations of @xmath80 and @xmath81 @xcite ,
imply that @xmath104 is vanishing everywhere except possibly at the shell .
the rightmost expression in ( [ pisfr - def ] ) shows that @xmath104 can not contain a delta - function at the shell , and @xmath104 is therefore everywhere vanishing . from now on
, we can therefore proceed assuming that @xmath93 and @xmath104 are continuous , and that their @xmath27-derivatives have at most finite discontinuities at isolated values of @xmath27 . by
( [ pim - def ] ) , the same will then hold for @xmath100 , with the exception of the horizon where @xmath100 diverges .
this tightens the neighborhood of the classical solutions in which the fields can take values , but it will not affect the critical points of the action .
the reason for this assumption is that it will make the terms @xmath105 and @xmath106 in our new action distributionally well defined .
we can now proceed to the liouville forms .
a direct computation yields @xmath107 \ \ .
\label{diff - liouville}\end{aligned}\ ] ] the variation @xmath15 affects the smoothness of the gravitational variables in the same way as the time derivative in subsection [ subsec : local - eoms ] . away from the horizon
parallel to the shell history , on which @xmath93 vanishes and @xmath108 diverges , all the terms in ( [ diff - liouville ] ) are therefore distributionally well defined : the terms on the left - hand side are at most discontinuous in @xmath27 , while the terms on the right - hand side may contain at worst delta - functions arising from @xmath109 .
the status at the horizon on which @xmath93 vanishes will be discussed below . to obtain the difference in the prospective liouville forms , we need to integrate the relation ( [ diff - liouville ] ) over @xmath27 . in an integral over a finite interval in @xmath27 ,
the only subtlety arises from the horizon on which @xmath93 vanishes . on a classical solution with mass @xmath19 ,
it can be shown from the embedding analysis of appendix [ app : static - shell ] that @xmath110 where the subscript @xmath111 indicates the values of the quantities at the horizon on which @xmath93 vanishes .
equations ( [ fminuseta - hor - fall ] ) and ( [ pim - def ] ) therefore show that , on the classical solution , the integral of ( [ diff - liouville ] ) across the horizon is well defined in the principal value sense , just as in the corresponding analysis of ref .
@xcite . to extend this argument off the classical solutions ,
we note that when the constraints hold , @xmath4 is constant in @xmath27 across this horizon .
as our action contains the constraints with their associated lagrange multiplies , we argue that @xmath4 can be assumed smooth at the horizon in the relation ( [ diff - liouville ] ) .
we can then again employ ( [ fminuseta - hor - fall ] ) and ( [ pim - def ] ) , and it is seen as above that the integral of ( [ diff - liouville ] ) across this horizon is well defined in the principal value sense .
what needs more attention is the falloff in ( [ diff - liouville ] ) at the infinities . from ( [ metricfall ] ) , ( [ trans ] ) , and ( [ pi - def ] ) , we have @xmath112 this means that the integrals of the third term on the left - hand side and the total variation term on the right - hand side diverge as @xmath66 . the geometrical reason for this divergence
is , as seen from ( [ pimprime ] ) , that @xmath100 is associated with a _
null _ time , rather than an asymptotically minkowski time .
the cure is to introduce convergence functions that provide the necessary translation between asymptotically spacelike hypersurfaces and asymptotically null hypersurfaces . to this end , let @xmath113 be a function that is smooth in @xmath27 and depends on our variables only through @xmath114 and @xmath115 as indicated .
let @xmath116 have the falloff @xmath117 adding @xmath118 on both sides of ( [ diff - liouville ] ) yields now an equation whose both sides can be integrated in @xmath27 from @xmath119 to @xmath120 .
the substitution terms arising from the first term on the right - hand side vanish , and we obtain @xmath121 dr \right\ } \ \ .
\label{int - diff - liouville}\end{aligned}\ ] ] all the terms in ( [ int - diff - liouville ] ) are well defined , provided the integral across the horizon on which @xmath93 vanishes is interpreted in the principal value sense .
we therefore see that the set @xmath122 provides a new canonical chart on the phase space .
note that this canonical transformation leaves the shell variables @xmath123 entirely invariant .
the geometrical meaning of the convergence function @xmath116 will be discussed in section [ sec : reduction ] .
what remains is to write the constraint terms in the action in terms of the new variables .
consider first the constraints away from the shell .
a straightforward rearrangement yields @xmath124 where @xmath125 note that @xmath126 is the same as in ref .
@xcite . both terms on the right - hand side of ( [ smooth - new - constraints ] ) are distributionally well defined . away from the shell , we can therefore include the constraints in the action in the form shown on the right - hand side of ( [ smooth - new - constraints ] ) , with @xmath127 and @xmath126 as independent lagrange multipliers .
this constraint redefinition is mildly singular on the horizon parallel to the shell history , owing to the divergence of @xmath100 ; however , one can argue as in ref.@xcite that the redefined constraints are equivalent to the old ones by continuity .
the falloff of the new multipliers is @xmath128 to recover the delta - constraint ( [ delta - constraints2-f ] ) , we observe from ( [ pisfr - def ] ) that ( [ delta - constraints2-f ] ) is equivalent to @xmath104 not having a delta - contribution at the shell .
we therefore argue that including in the action the constraint term @xmath129 , with @xmath126 an independent lagrange multiplier , yields both the constraint @xmath130 away from the shell and the delta - constraint ( [ delta - constraints2-f ] ) at the shell .
need not be continuous at the shell , not even on the classical solutions
. the product @xmath131 would therefore not be distributionally well defined in the event that @xmath104 did contain a delta - contribution at the shell . ] finally , consider the delta - constraint ( [ delta - constraints2-pl ] ) . using ( [ pim - def ] ) and ( [ delta - constraints2-f ] ) , equation ( [ trans - m ] ) implies @xmath132 when ( [ delta - constraints2-f ] ) holds , ( [ delta - constraints2-pl ] ) is thus equivalent to @xmath133 including in the action the constraint term @xmath134 $ ] , with @xmath127 an independent lagrange multiplier , therefore yields both the constraint @xmath135 away from the shell and the delta - constraint ( [ delta - constraints2-pl ] ) at the shell .
these considerations have led us to the action @xmath136 \nonumber \\ & & - \int dt \ ; \int_{-\infty}^\infty dr \left\ { n^\sfr \pi_\sfr + { \tilde n } [ m ' - \zp \pi_m^{-1 } \delta(r-\zr ) ] \right\ } \nonumber \\ & & -\int dt \ , \left ( n_+m_+ + n_- m_- \right ) \ \ .
\label{new - action - mpimdot}\end{aligned}\ ] ] both the action and its variations are well defined .
the poisson bracket algebra of the constraints clearly closes .
note that the convergence term @xmath137 in no way contributes to the local variations of the action .
the liouville term @xmath138 can be brought to a form in which the time derivative is on @xmath19 , at the cost of introducing another convergence term .
let @xmath139 be a smooth function of @xmath27 only , with the falloff @xmath140 we then have @xmath141 all the terms in ( [ mpim - int ] ) are well defined , and each side can be integrated in @xmath27 from @xmath119 to @xmath120 .
we arrive at the action @xmath142 \right\ } \nonumber \\ & & - \int dt \ ; \int_{-\infty}^\infty dr \left\
{ n^\sfr \pi_\sfr + { \tilde n } [ m ' - \zp \pi_m^{-1 } \delta(r-\zr ) ] \right\ } \nonumber \\ & & -\int dt \ , \left ( n_+m_+ + n_- m_- \right ) \ \ .
\label{new - action - pimmdot}\end{aligned}\ ] ] the geometrical meaning of the convergence terms will become explicit in section [ sec : reduction ] .
in subsection [ subsec : trans - nonhorizon ] we excluded the special case where the shell straddles a horizon .
we now discuss how this special case can be included .
when the shell straddles a horizon , the zero of @xmath93 occurs at the shell .
the delta - constraints at the shell are given by ( [ delta - constraints2 ] ) .
when the equations of motion hold , the masses on the two sides agree , and the embedding analysis of appendix [ app : static - shell ] shows that equation ( [ fminuseta - hor - fall ] ) holds , now with @xmath143 and @xmath144 . from ( [ pisfr - def ] )
we see that @xmath104 can not contain a delta - function at the shell .
we can therefore again assume that @xmath93 and @xmath104 are continuous , and that their @xmath27-derivatives have at most finite discontinuities at isolated values of @xmath27 .
the new feature in equation ( [ diff - liouville ] ) is that the singularity of @xmath145 and @xmath100 now occurs at the shell .
when the equations of motion hold , we see from ( [ pim - def ] ) and ( [ fminuseta - hor - fall ] ) that integrating each side of ( [ diff - liouville ] ) in @xmath27 across the shell is well defined in the principal value sense , and we argue as above that this conclusion can be extended away from the classical solutions provided the constraints are understood to hold .
we argue similarly that the left - hand side of ( [ diff - liouville ] ) contains no delta - contributions at the shell , and it is therefore justified to interpret the integral of ( [ diff - liouville ] ) over @xmath27 as the principal value .
convergence at the infinities is accomplished as above , and the substitution terms from the total @xmath27-derivative on the right - hand side of ( [ diff - liouville ] ) vanish .
we therefore again arrive at ( [ int - diff - liouville ] )
. equations ( [ deltam1 ] ) and ( [ deltam2 ] ) remain valid , with the understanding @xmath146 , and the delta - constraints can be taken in the action as before . to justify the manipulations leading to the action ( [ new - action - pimmdot ] )
, we again appeal to the constraints to argue that @xmath19 can be regarded as smooth in @xmath27 at the shell , and that @xmath147 then does not contain a delta - function at the shell .
we therefore see that the actions ( [ new - action - mpimdot ] ) and ( [ new - action - pimmdot ] ) remain valid in a suitable limiting sense also for a horizon - straddling shell .
in this section we eliminate the constraints and find the dynamics in the reduced phase space . we shall continue to treat the cases @xmath46 separately , and we denote the corresponding two components of the reduced phase space by @xmath148 .
we first assume , in subsection [ subsec : reduction - not - hor ] , that the shell history does not lie on a horizon , and we then include the horizon - straddling shell as a limiting case in subsection [ subsec : reduction - on - hor ] . finally , in subsection [ subsec : reduction - no - shell ] , we attach the right - moving and left - moving test shell limits to the respective components of the reduced phase space as regular boundaries .
it will be useful in the reduction to assume a more definite form for the convergence function @xmath116 . from now on , we take [ g - special ] @xmath149 where @xmath150 are smooth functions of @xmath27 only , with the falloff @xmath151 where @xmath152 denotes the step function . in this subsection
we assume that the shell history does not lie on a horizon . solving the constraint
@xmath130 implies that @xmath87 and @xmath104 simply drop out of the action . to solve the remaining constraint , @xmath153 we write @xmath154 where @xmath155 are regarded as independent variables .
we then have @xmath156 and the constraint ( [ red - constr - pim ] ) implies @xmath157 note that as the shell history does not lie on a horizon , each of the two factors on the right - hand side of ( [ zp - sol - red ] ) is nonvanishing . using ( [ g - special - fall ] ) , ( [ dotm - red ] ) , and ( [ zp - sol - red ] ) , we find @xmath158 & = & \bpplus \dot{\bm}_+ + \bpminus \dot{\bm}_- \nonumber \\ & & + { d\over dt } \left [ \eta(\bmplus-\bmminus ) \int_0^\zr g \ , dr \right ] \ \ , \label{lioville - reduction}\end{aligned}\ ] ] where [ bmpplusminus ] @xmath159 \ \ , \label{bmpplus } \\ \bpminus & : = & \int_{-\infty}^\infty dr \left [ \pi_m \theta(\zr - r ) - \eta g \theta(-r ) - 2\eta \bmminus g_- \right ] \ \ . \label{bmpminus}\end{aligned}\ ] ] the singularity of @xmath160 occurs in precisely one of the two integrals in ( [ bmpplusminus ] ) , and the integral over this singularity is interpreted in the principal value sense . substituting ( [ lioville - reduction ] ) into ( [ new - action - pimmdot ] ) , and dropping the integral of a total derivative , we obtain the reduced action @xmath161 this shows that the set @xmath162 provides local canonical coordinates on @xmath148 .
the equations of motion derived from the action ( [ red - action1 ] ) read @xmath163 the emergence of @xmath7 as two coordinates on @xmath148 is not surprising : on a classical solution , @xmath7 are the two schwarzschild masses , and these masses together with @xmath164 completely determine the four - dimensional spacetime .
to understand the geometrical meaning of @xmath8 , we recall from ( [ pimprime ] ) that without the convergence terms proportional to @xmath165 and @xmath166 , the integrals in ( [ bmpplusminus ] ) would give the time differences between the shell and the infinities on the constant @xmath26 hypersurface .
as the constant @xmath26 hypersurface extends to the spacelike infinities , such null - time differences would be infinite .
the role of the convergence terms in ( [ bmpplusminus ] ) is to absorb the infinities : one can think of the convergence terms as associating to the constant @xmath26 hypersurface a hypersurface that is asymptotically _ null _ as @xmath66 . for @xmath62 ,
this associated hypersurface extends from the left - hand - side @xmath167 to the right - hand - side @xmath168 ( points @xmath169 and @xmath170 in figure [ fig : nonhorizon ] ) ; for @xmath61 , the situation is the reverse .
thus , @xmath171 is the time difference between the shell and the right - hand - side infinity of the associated asymptotically null hypersurface , and @xmath172 is the time difference between the shell and the left - hand - side infinity of the associated asymptotically null hypersurface .
the equations of motion ( [ dotbpplusminus ] ) show that the time evolution of @xmath8 only arises from the evolution of the constant @xmath26 hypersurfaces at the infinities .
thus , in this local canonical chart on @xmath148 , the information about the shell motion is encoded in equations ( [ dotbpplusminus ] ) .
it should be emphasized that the degrees of freedom present in @xmath8 are invariant under the isometries of the spacetime . killing time translations on the spacetime move both the shell history and the constant @xmath26 hypersurface
: in particular , they move the two asymptotic ends of the constant @xmath26 hypersurface , and hence the asymptotic ends of the associated asymptotically null hypersurface
. however , the time _ differences _ that constitute the momenta are invariant under killing time translations .
as we have assumed that the shell history does not lie on a horizon , the coordinates @xmath162 do not form a global chart on @xmath148 .
instead , these coordinates provide two disjoint local canonical charts , covering two disconnected sets in @xmath148 : one for @xmath173 and the other for @xmath174 , with unrestricted values of @xmath8 in each chart .
these coordinates can not be extended to @xmath175 .
the reason is that when @xmath175 , the shell history lies on a horizon , the singularity in @xmath100 is at @xmath176 , and the first term under each integral in ( [ bmpplusminus ] ) makes both @xmath171 and @xmath177 divergent .
we shall address the special case @xmath175 and the global structure of @xmath148 in subsection [ subsec : reduction - on - hor ] . it will be useful to introduce on @xmath148 another set of local canonical coordinates , @xmath178 , by the transformation [ plusminus - to - tildebars ] @xmath179 the inverse transformation is @xmath180 from ( [ bmpplusminus ] ) , ( [ bpbar - def ] ) , and ( [ bptilde - def ] ) , we see that @xmath181 contains the information about the asymptotic ends of the constant @xmath26 hypersurface , whereas the information about the location of the shell with respect to the infinities is encoded in @xmath182 .
we can therefore loosely regard the pair @xmath183 as describing the vacuum spacetime dynamics , and the pair @xmath184 as describing the shell . while not literal , this view will be helpful for understanding the global structure of @xmath148 in subsections [ subsec : reduction - on - hor ] and [ subsec : reduction - no - shell ] .
as defined by the transformation ( [ plusminus - to - tildebars ] ) , the coordinates @xmath178 provide two disjoint local canonical charts that cover on @xmath148 the same two disconnected sets as the coordinates @xmath162 .
the ranges of the variables in these two charts are respectively @xmath185 and @xmath186 , each with unrestricted @xmath181 and @xmath182 .
the coordinates @xmath187 can not , however , be extended to @xmath188 .
while @xmath181 remains finite for a horizon - straddling shell , it is seen from ( [ bmpplusminus ] ) that @xmath182 must diverge .
we now wish to find on @xmath148 coordinates that extend to the horizon - straddling shell .
we shall first rely on the spacetime picture to identify the geometrical information that the coordinates must carry in this limit .
we then construct on @xmath148 a global canonical chart that contains this information .
consider the spacetime of figure [ fig : nonhorizon ] .
the shell is left - moving , corresponding to @xmath62 , and the shell history lies in the future of the left - going horizon , corresponding to @xmath189 .
the points @xmath169 and @xmath170 indicate the ends of the asymptotically null hypersurface that is associated to the hypersurface of constant @xmath26 .
@xmath171 is the difference in the time between points @xmath190 and @xmath191 , and @xmath177 is the difference in the time between points @xmath192 and @xmath193 . in this spacetime ,
let @xmath194 be the radial null geodesic connecting @xmath190 to @xmath191 , let @xmath195 the radial null geodesic connecting @xmath192 to @xmath193 , and let @xmath196 be the radial null geodesic connecting @xmath190 to @xmath193 .
let @xmath197 ( @xmath198 ) be the affine parameters on these geodesics , each normalized to have the range @xmath199 $ ] .
@xmath200 and @xmath201 increase toward the future .
increases toward the future if @xmath191 is in the future of @xmath190 as shown in the figure , corresponding to @xmath203 , and it increases towards the past if @xmath191 is in the past of @xmath190 , corresponding to @xmath204 . in the special case @xmath205 , @xmath191 and @xmath190 coincide , and @xmath194 degenerates to a point .
we now define the quantities @xmath206 by [ q12-def ] @xmath207 similarly , consider a spacetime in which @xmath62 but the shell history lies in the past of the left - going horizon , corresponding to @xmath208 . in this spacetime ,
the counterparts of points @xmath190 and @xmath193 are below the left - going horizon , but the three null geodesics @xmath209 ( @xmath198 ) can be defined as above , the only modification being that the potentially degenerate one is now @xmath195 . in this spacetime , we again define @xmath206 by ( [ q12-def ] ) . a straightforward calculation yields [ qpm - ab - bel ] @xmath210 \ \ , \\ \noalign{\smallskip } q_- & = & -8 \bmminus \left [ \bmplus - \bmminus + |\bmplus-\bmminus| \exp\left(\frac{\bpminus}{4\bmminus}\right ) \right ] \
\ , \end{aligned}\ ] ] valid both for @xmath189 and @xmath208 .
the canonical coordinates @xmath162 can therefore be replaced by the noncanonical coordinates @xmath211 , with @xmath212 for @xmath189 and @xmath213 for @xmath208 .
the crucial observation is now that @xmath206 , as defined in ( [ q12-def ] ) , remain well defined also for the spacetimes shown in figure [ fig : horizon ] , in which @xmath175 and the shell history lies on a common horizon . in these spacetimes
, @xmath206 can each take arbitrary negative values . as the soldering along the horizon is affine , @xmath206
precisely encode the coordinate - invariant information about the relative loci of the points @xmath190 , @xmath193 , @xmath169 , and @xmath170 ( or , equivalently , the points @xmath190 , @xmath193 , @xmath191 , and @xmath192 ) .
this means that the set @xmath214 provides a global , noncanonical chart on @xmath215 .
the domain is @xmath212 for @xmath189 , @xmath213 for @xmath208 , and @xmath216 for @xmath175 . in the above construction
we have taken @xmath62 .
it is clear that an entirely analogous discussion carries through for @xmath61 , with straightforward changes in formulas ( [ qpm - ab - bel ] ) , and yielding a global , noncanonical chart on on @xmath217 . to find a global _ canonical _ chart on @xmath148 , consider the transformation [ bpi - def ] @xmath218 equations ( [ bpi - def ] ) clearly define a canonical transformation from @xmath219 to @xmath220 individually in the domains @xmath185 and @xmath186 .
it is straightforward to verify that the chart @xmath220 becomes global on @xmath148 when extended to @xmath188 with unrestricted values of @xmath221 and @xmath222 . for @xmath62 , in particular , ( [ qpm - ab - bel ] ) shows that @xmath206 can be written as @xmath223 -\bmtilde \right\ } \ \ , \\
\noalign{\smallskip } q_- & = & -16 ( \bmbar - \bmtilde ) \left\ { \bmbar \exp \ ! \left [ - \left ( \frac{\bmbar + \bmtilde}{\bmbar - \bmtilde } \right ) + \frac{\bpibar - \bpitilde}{8(\bmbar - \bmtilde ) } \right ] + \bmtilde \right\ } \ \ , \end{aligned}\ ] ] from which the regularity of the @xmath224 limit is manifest .
we have thus shown that the set @xmath220 provides a global canonical chart on @xmath148 .
the domain of the variables is @xmath225 , with @xmath221 and @xmath222 taking all real values .
we therefore have @xmath226 .
the hamiltonian reads @xmath227 the values of @xmath228 and @xmath229 are constants of motion , whereas the equations of motion for @xmath221 and @xmath222 show that the evolution of @xmath206 only arises from the evolution of the constant @xmath26 hypersurface at the two spacelike infinities .
this means that the information about the shell dynamics is contained in the momentum equations of motion both for @xmath230 ( as was already seen in subsection [ subsec : reduction - not - hor ] ) as well as for @xmath188 .
we have so far assumed that the unreduced shell momentum @xmath44 is nonvanishing .
we saw that this assumption is compatible with the dynamics , and that it divides the reduced phase space into the two disconnected sectors @xmath148 , labeled by @xmath231 . as the unreduced bulk action ( [ s - ham ] )
is not differentiable in @xmath44 at @xmath45 , it is not clear whether @xmath217 and @xmath215 are joinable to each other in any smooth sense .
our reduction formalism is not well suited to examining this issue : the canonical transformation of section [ sec : transformation ] was tailored to the null hypersurfaces separately for @xmath46 .
we can , however , address the limit @xmath232 _ individually _ in @xmath217 and @xmath215 . as the shell stress - energy tensor ( [ shell - tab ] )
vanishes for @xmath232 , this is the limit of a test shell that traverses the spacetime without affecting it gravitationally .
we shall now show that one can attach the right - moving and left - moving test shell limits respectively to @xmath217 and @xmath215 as smooth boundaries with topology @xmath2 .
when the test shell history does not lie on a horizon , the situation is straightforward .
we can start with the coordinates @xmath162 , separately for @xmath173 and @xmath174 , and simply take the limit @xmath175 with @xmath8 remaining finite . from the geometrical interpretation of @xmath8
it is seen that this attaches to @xmath148 those test shell configurations in which the test shell does not straddle a horizon .
the locus of the test shell history is determined by @xmath8 exactly as in subsection [ subsec : reduction - not - hor ] .
including a horizon - straddling test shell is more intricate . in the global chart @xmath220 ,
the limit of a test shell on a horizon is achieved by setting first @xmath188 and then taking @xmath233 while keeping @xmath221 finite .
on the other hand , the limit of a test shell off the horizon requires taking simultaneously @xmath224 and @xmath233 so that @xmath181 and @xmath182 remain finite .
what we need is a new canonical chart in which both of these limits are brought to finite values of the coordinates . to this end , let [ bd - chart ] @xmath234 as @xmath235 equations ( [ bd - chart ] ) define a canonical transformation from the chart @xmath220 to the chart @xmath236 .
the new canonical chart is global : the range is @xmath237 and @xmath238 , with unrestricted @xmath239 and @xmath240 . the qualitative location of the shell history is governed by the sign of @xmath240 : @xmath241 ( @xmath242 ) yields a shell in the future ( past ) of the horizon that is parallel to the shell history , while @xmath243 yields a shell history on the horizon .
it is now easily seen that in this chart the test shell limit is @xmath244 , with the other coordinates remaining at finite values .
a horizon - straddling test shell is recovered with @xmath243 , whereas @xmath241 ( @xmath242 ) gives a test shell in the future ( past ) of the horizon that the test shell does not cross .
clearly , the test shell limit constitutes a smooth boundary of @xmath148 with topology @xmath2 .
while the charts on @xmath148 introduced in section [ sec : reduction ] are well adapted to the geometry of the spacetime , they contain the information about the shell motion in a nontransparent manner . in this section
we introduce on @xmath148 a local canonical chart that describes more directly the motion of the shell in the spacetime geometry .
a chart of this kind is of particular physical interest if one wishes to quantize the system as a model of black hole radiation with back reaction @xcite .
the physical situation we have in mind is a static observer who scrutinizes the shell motion from an asymptotically flat infinity . for definiteness
, we take this infinity to be the right - hand - side one .
we set @xmath245 , so that the coordinate time @xmath26 coincides with the observer s proper time . to incorporate the observer s ignorance of what is happening at the left - hand - side infinity , we set @xmath246 .
we further assume that the shell history reaches a future or past null infinity on the right - hand side .
the penrose diagram for @xmath62 is therefore as in figure [ fig : nonhorizon ] , and the penrose diagram for @xmath61 is the time inverse . in particular , we have @xmath247 , and we are in the region of @xmath148 covered by the chart @xmath6 with @xmath173 .
consider thus the chart @xmath6 with @xmath173.from section [ sec : reduction ] we recall that the pair @xmath11 only carries information about the geometry right of the shell , and the pair @xmath248 only carries information about the geometry left of the shell . to describe the motion of the shell as seen from the right - hand - side infinity
, we can therefore leave the pair @xmath248 intact and seek a canonical transformation that replaces @xmath11 by a new pair . to specify the new pair , we choose in the kruskal geometry right of the shell a stationary coordinate system that conforms to the falloff ( [ metricfall ] ) with @xmath245 . as @xmath249 is then an increasing function , we can assume @xmath250 without loss of generality .
let @xmath251 stand for the shell curvature radius in these coordinates : @xmath252 .
we now seek a momentum @xmath253 such that there is a canonical transformation from the pair @xmath11 to the pair @xmath254 .
writing the transformation as @xmath255 and @xmath256 , the canonicality criterion reads @xmath257 \over \partial \bmplus } = - { \partial [ \bpplus(\rhat , \bmplus ) ] \over \partial \rhat } \ \ .
\label{can - crit}\ ] ] substituting @xmath171 from ( [ bmpplus ] ) to the right - hand side of ( [ can - crit ] ) yields @xmath257 \over \partial \bmplus } = \pi_m(\rhat , \bmplus ) \ \ , \label{phatvpi}\ ] ] where @xmath258 is determined by the choice of the stationary coordinate system .
note that the convergence functions @xmath139 and @xmath259 have not entered ( [ phatvpi ] ) . solving the differential equation ( [ phatvpi ] ) for @xmath260 yields the desired canonical transformation , and inverting this solution gives @xmath261 as a function in the new canonical chart .
the action reads @xmath262 \ \ .\ ] ] the shell stress - energy tensor in the new chart can be found using ( [ ell - def ] ) , ( [ shell - tab ] ) and ( [ zp - sol - red ] ) . as explicit examples , we now present the shell
hamiltonians @xmath261 in four different stationary coordinate systems .
we arrived at the first three coordinate systems by seeking a simple functional form for @xmath261 .
the fourth coordinate system is the spatially flat one used in ref . @xcite . as a first example , we consider coordinates in which the metric reads [ polygauge - metric ] @xmath263 with @xmath264 , these coordinates cover half of kruskal manifold in the appropriate manner , and the falloff ( [ metricfall ] ) is satisfied with @xmath265 .
the relation to the curvature coordinates is @xmath266 we find @xmath267 solving ( [ phatvpi ] ) with a convenient choice for the integration constant , we obtain @xmath268 the equation of motion derived from the hamiltonian ( [ polygauge - ham ] ) can be integrated as @xmath269 .
it is easily verified that this is the correct equation for a null geodesic in the metric ( [ polygauge - metric ] ) .
consider next coordinates in which the metric reads @xmath270 \ \ , \\ & & \lambda^2 n^r = - \eta \left [ 1 - ( 1 - 2\bmplus / r)\exp(2\bmplus / r ) \right ]
\ \ .\end{aligned}\ ] ] with @xmath264 , these coordinates cover half of kruskal manifold in the appropriate manner , and the falloff ( [ metricfall ] ) is satisfied with @xmath265 .
the relation to the curvature coordinates is @xmath271 dr '
\ \ .\end{aligned}\ ] ] we find @xmath272 solving ( [ phatvpi ] ) with a convenient choice for the integration constant , we obtain @xmath273 the equation of motion can be solved implicitly in terms of the exponential integral function .
consider next coordinates in which the metric reads [ eddfinkgauge - metric ] @xmath274 with @xmath264 , these coordinates cover half of kruskal manifold in the appropriate manner .
the relation to the curvature coordinates is @xmath275 we recognize these coordinates as simply related to the coordinates @xcite : @xmath276 is the retarded ( advanced ) time for @xmath61 ( @xmath62 ) . in terms of the tortoise
coordinate @xmath277 $ ] , we have @xmath278 .
there is a minor technical issue in that the coordinates ( [ eddfinkgauge - metric ] ) do not obey the falloff ( [ metricfall ] ) : we have @xmath279 , which violates ( [ plambdafall ] ) .
we therefore take the coordinates ( [ eddfinkgauge - metric ] ) to hold for @xmath280 , where @xmath281 is a large parameter , and smoothly deform them to a faster falloff for @xmath282 . as equation ( [ phatvpi ] )
is local in @xmath251 , the form of the canonical transformation for @xmath280 is independent of that for @xmath282 . in the end , we can either leave the hamiltonian unspecified for @xmath282 , or argue that one can take the limit @xmath283 in the sense of some suitable renormalization in the parameters of the canonical transformation . proceeding in this way
, we find @xmath284 and , with a convenient choice of the integration constant , @xmath285 the hamiltonian ( [ efgauge - ham ] ) clearly correctly reproduces the fact that the time @xmath276 is constant on the shell history . as the last example
, we consider coordinates in which the metric reads [ spatflatgauge - metric ] @xmath286 with @xmath264 , these coordinates cover half of kruskal manifold in the appropriate manner .
the relation to the curvature coordinates is @xmath287 we recognize these coordinates as the spatially flat coordinates @xcite , recently employed in the study of hawking radiation with back reaction in ref .
@xcite .
there is again a minor technical issue in that the coordinates ( [ spatflatgauge - metric ] ) do not obey the falloff ( [ metricfall ] ) . a hamiltonian falloff analysis compatible with these coordinates has been discussed in the metric variables in ref .
@xcite . here
, however , we shall simply argue in terms of a cutoff parameter @xmath281 as above .
we have @xmath288 and solving ( [ phatvpi ] ) with a suitable integration constant yields @xmath261 implicitly as the solution to @xmath289 in order to make a connection to the work in ref .
@xcite , we define [ kw - momenta ] @xmath290 \ \ , \label{kw - momenta - pc } \\
\bp_c & : = & \bpminus - \eta \left [ \rhat - 2 \sqrt{2\bmminus\rhat } + 4 \bmminus \ln \left ( 1 + \sqrt{\rhat/(2\bmminus ) } \right ) \right ] \ \ .\end{aligned}\ ] ] equations ( [ kw - momenta ] ) define a canonical transformation from the chart @xmath291 to the new canonical chart @xmath292 .
the hamiltonian @xmath293 in the new chart is obtained ( in implicit form ) by eliminating @xmath253 from ( [ bmplus - implicit ] ) and ( [ kw - momenta - pc ] ) .
as the value of @xmath294 is a constant of motion , the system can be partially reduced by regarding @xmath294 as a prescribed constant .
the term @xmath295 then drops from the action , and we obtain @xmath296 \ \ .\ ] ] for @xmath62 , this is the action derived in refs.@xcite by different methods . for @xmath61 , it is not .
the reason is that the coordinates ( [ spatflatgauge - metric ] ) are the ingoing spatially flat coordinates for @xmath62 and the outgoing spatially flat coordinates for @xmath61 , thus covering all of the spacetime right of the shell in each case , whereas ref.@xcite was physically motivated to use the ingoing spatially flat coordinates irrespectively the direction of the shell motion
. it would be straightforward to repeat the above analysis with the sign of @xmath164 in ( [ spatflatgauge - metric ] ) reversed , recovering the result of refs .
@xcite for @xmath61 .
note , however , that with the sign of @xmath164 in ( [ spatflatgauge - metric ] ) reversed , the coordinates do not cover the part of the shell history that lies inside the horizon .
in the previous sections we fixed the evolution of the spatial hypersurfaces at the spacelike infinities by taking @xmath69 to be prescribed functions of @xmath26 . in this section
we free this evolution by making the replacement @xcite @xmath297 in the boundary term in the actions ( [ total - metric - action ] ) and ( [ new - action - pimmdot ] ) .
the variations of @xmath23 become then unrestricted at @xmath66 , but varying the action with respect to @xmath298 yields the relations ( [ n - vs - tau ] ) as equations of motion .
the new variables @xmath298 are the proper times measured by static standard clocks at the respective infinities , with the convention that @xmath299 increases toward the future and @xmath300 increases toward the past . in the absence of a shell
, it was shown in ref .
@xcite that the action containing @xmath298 as independent variables can be brought to a canonical form in which the unconstrained degrees of freedom and the pure gauge degrees of freedom are entirely decoupled .
we now outline the analogous result in the presence of the null shell . for brevity , we shall refrain from explicitly spelling out the smoothness properties of the various emerging phase space functions .
we start from the action ( [ new - action - pimmdot ] ) , with @xmath116 given by ( [ g - special ] ) , and we make in the boundary term the replacement ( [ n - vs - tau ] ) .
the resulting action is a sum of two decoupled parts : a hamiltonian action @xmath301 consisting of the terms that contain the pair @xmath302 , and the remainder @xmath303 .
we only need to consider @xmath303 . as in section
[ sec : transformation ] , we assume first that the shell history does not lie on a horizon , and relax this assumption at the end . under the time integral in @xmath303 , the terms homogeneous in the time derivatives are @xmath304 \ \ .\ ] ] we pass from the noncanonical chart @xmath305 to the new chart @xmath306 , defined by @xmath307 \left [ \theta(r'-r ) - \theta(r - r ' ) \right ] \nonumber \\ & & \qquad\qquad\qquad\qquad\qquad\quad - \eta m_+ g_+(r ' ) + \eta m_- g_-(r ' ) \biggr\ } \ \ , \label{pigamma } \\
\mbar & : = & \casehalf ( m_+ + m_- ) \ \ , \label{mbardef } \\ \pbar & : = & \tau_+ - \tau_- + \int_{-\infty}^\infty dr \ , \biggl\ { [ \pi_m(r ) - \eta g(r ) ] - 2\eta m_+ g_+(r ) - 2\eta m_- g_-(r ) \biggr\ } \ \ .\end{aligned}\ ] ] the falloff is [ gammapigammafall ] @xmath308 by techniques similar to those in ref .
@xcite , we find @xmath309 the chart @xmath306 is therefore canonical . dropping the integral of a total derivative , the action reads @xmath310 \ \ .
\label{snought - can1}\end{aligned}\ ] ] to express the constraint in the new chart , we have used equation ( [ gamma ] ) and the relation @xmath311 , which follows by differentiating ( [ pigamma ] ) . the action ( [ snought - can1 ] ) is canonical , but the constraint couples the variables in a nontransparent way . to decouple the degrees of freedom ,
we pass to the chart @xmath312 , defined by @xmath313}^{-1 } \delta(r-\zr ) \ \ , \label{tildegamma } \\
\pi_{\tilde\gamma}(r ) & : = & \pi_\gamma(r ) \ \ , \\
\mtilde & : = & \casehalf \zp { \left(\eta \hat{g } - \widehat{\pi'_{\gamma } } \right)}^{-1 } \ \ , \label{mtildedef } \\ \ptilde & : = & 2 \widehat{\pi_\gamma } - 2 \eta \int_0^\zr dr \ , g(r ) \ \ .\end{aligned}\ ] ] the falloff of @xmath314 and @xmath315 is clearly the same as that of @xmath316 and @xmath317 , given in ( [ gammapigammafall ] ) . using the analogue of relation ( [ total - der - rhat ] ) for @xmath317
, we find @xmath318 \ \ .\ ] ] the chart @xmath312 is therefore canonical .
dropping the integral of a total derivative , the action reads @xmath319 the unconstrained canonical degrees of freedom , @xmath320 , have now become decoupled from the pure gauge degrees of freedom . to put the action in a more transparent form ,
we write @xmath321 \ \ , \label{vrdef } \\
\pi_v(r ) & : = & - \tilde{\gamma}(r ) \ \ .\end{aligned}\ ] ] the falloff is [ vpvfall ] @xmath322 as @xmath323 the transformation to the chart @xmath324 is not canonical as it stands .
however , it becomes canonical after the first term on the right - hand side of ( [ dotpbarpiv ] ) is absorbed into the constraint term by writing @xmath325 and regarding @xmath326 as a new lagrange multiplier .
as the equations of motion imply @xmath327 , the falloff of @xmath326 is @xmath328 dropping the integral of a total derivative , and including @xmath301 , we finally obtain the action @xmath329 all the variables in the action ( [ fheis - action ] ) have a transparent geometrical meaning . from ( [ gamma ] ) , ( [ mbardef ] ) , ( [ tildegamma ] ) , and ( [ mtildedef ] ) , we see that @xmath330 and @xmath331 are respectively equal to the variables @xmath228 and @xmath229 introduced in section [ sec : reduction ] .
similarly , using ( [ n - vs - tau ] ) , we see that @xmath332 and @xmath333 can be interpreted as the time - independent initial values of the variables @xmath181 and @xmath182 introduced in section [ sec : reduction ] .
as for the pure gauge degrees of freedom , @xmath87 is the curvature radius , and equation ( [ vrdef ] ) shows that @xmath334 is the time .
the action ( [ fheis - action ] ) therefore provides a natural generalization of the vacuum action given in eq .
( 149 ) of ref .
@xcite .
we have here assumed that the shell history does not lie on a horizon .
this assumption can be relaxed , in a suitable limiting sense , by performing on the coordinates @xmath320 transformations analogous to those given for the coordinates @xmath219 in section [ sec : reduction ] .
in this section we consider the canonical transformation and hamiltonian reduction for spatial topology @xmath2 . for concreteness , we take the evolution of the spatial hypersurfaces at the single spacelike infinity to be prescribed as in section [ sec : metric ]
. it will be seen that the reduced phase space consists of two disconnected components , one for an expanding shell and the other for a collapsing shell .
each component has the topology @xmath10 .
we start from the action principle . in the bulk action ( [ s - ham ] ) , we take @xmath335 , with the falloff ( [ metricfall ] ) as @xmath336 . as @xmath337 , we introduce the falloff [ s - metricfall ] @xmath338 where @xmath339 , @xmath340 , @xmath341 , @xmath342 , @xmath343 , and @xmath344 are functions of @xmath26 only .
it is straightforward to verify that the falloff ( [ s - metricfall ] ) is consistent with the constraints and preserved by the time evolution . by
( [ trans - m ] ) and ( [ f - def ] ) , the falloff ( [ s - metricfall ] ) implies that the mass left of the shell must vanish when the equations of motion hold : @xmath345 is then just the coordinate singularity at the center of hyperspherical coordinates in flat space .
the classical solutions therefore describe a shell with a flat interior , and the spatial topology is @xmath2 .
the action appropriate for fixing @xmath346 is @xmath347 where @xmath348 is given by ( [ s - ham ] ) with @xmath335 .
the canonical transformation of section [ sec : transformation ] goes through with the obvious changes .
the new action is as in ( [ new - action - mpimdot ] ) , except that the integral is from @xmath345 to @xmath349 and the term @xmath350 is missing .
@xmath139 and @xmath351 are smooth in @xmath27 and have the same behavior as @xmath352 as in section [ sec : transformation ] .
the falloff of the new fields as @xmath337 can be found from ( [ s - metricfall ] ) ; for example , we have @xmath353 all the new fields remain regular as @xmath337 .
in particular , @xmath4 tends to zero as @xmath337 .
the hamiltonian reduction proceeds as in section [ sec : reduction ] , with the simplification that the interior mass vanishes .
the reduced phase space consists again of two disconnected components , denoted now by @xmath354 .
we take @xmath351 to be as in ( [ g - special ] ) with @xmath355 , and we solve the constraint ( [ red - constr - pim ] ) as in ( [ m - sol ] ) and ( [ zp - sol - red ] ) with @xmath356 . note that as @xmath100 has the same sign as @xmath44 , equation ( [ zp - sol - red ] ) implies @xmath357 .
we find @xmath358 = \bpplus \dot{\bm}_+ + { d\over dt } \left ( \eta \bmplus \int_0^\zr g \ , dr \right ) \ \ , \ ] ] where @xmath359 \ \ .
\label{bpplus - flat}\ ] ] substituting this in the action and dropping the integral of a total derivative , we obtain the reduced action @xmath360 thus , the pair @xmath11 provides a canonical chart on @xmath354 .
as @xmath100 does not have singularities , the definition ( [ bpplus - flat ] ) is always good : the chart is global , and the topology of @xmath354 is @xmath10 .
the test shell limit can be attached as a smooth boundary with topology @xmath361 at @xmath362 .
the information about the shell motion is again encoded in the evolution of @xmath171 .
charts that describe the shell motion in the exterior geometry more transparently can be constructed as in section [ sec : gauges ] .
in this paper we have analyzed the hamiltonian structure of spherically symmetric einstein gravity coupled to an infinitesimally thin null - dust shell .
we formulated the theory under kruskal - like boundary conditions , prescribing the evolution of the spatial hypersurfaces at the two spacelike infinities .
we adopted smoothness conditions that made the variational equations distributionally well defined , and equivalent to the einstein equations for this system .
we then simplified the constraints by a kucha - type canonical transformation and performed the hamiltonian reduction .
it was seen that the reduced phase space consists of two disconnected copies of @xmath1 , one for a right - moving shell and the other for a left - moving shell .
we found on each component a global canonical chart in which the configuration variables are the schwarzschild masses on the two sides of the shell , leaving the shell dynamics indirectly encoded in the conjugate momenta . excluding the special case of a shell straddling a horizon
, we found a local canonical chart in which the configuration variables are the shell curvature radius and the interior mass , in an arbitrarily specifiable stationary coordinate system exterior to the shell .
in particular , performing a partial reduction and fixing the interior mass to be a prescribed constant , we reproduced a previously known shell hamiltonian in the spatially flat gauge outside the shell .
we also cast into canonical form the theory in which the evolution at the infinities is freed by introducing parametrization clocks .
we found on the unreduced phase space a global canonical chart in which the physical degrees of freedom and the pure gauge degrees of freedom are completely decoupled , and we identified the pure gauge configuration variables in this chart as the time and the curvature radius .
finally , we adapted the analysis to the spatial topology @xmath2 , which has just one infinity , and for which the spacetime inside the shell is flat .
expectedly , the reduced phase space for this spatial topology turned out to consist of two disconnected copies of @xmath10 , one for an expanding shell and the other for a collapsing shell .
in addition to the kruskal spatial topology @xmath9 and the euclidean spatial topology @xmath2 , yet another spatial topology of interest would be that of the @xmath363 geon @xcite , @xmath364@xmath365a point at infinity@xmath366 . as the reduced phase space of the vacuum theory with the @xmath363 geon topology has dimension two @xcite
, one expects that the reduced phase space with a null shell would have dimension four .
indeed , this is the conclusion reached under a technically slightly different but qualitatively similar falloff in ref.@xcite , by first performing a hamiltonian reduction for a massive dust shell and then taking the zero rest mass limit .
it does not seem straightforward to adapt the canonical transformation of section [ sec : transformation ] to @xmath363 geon topology , however .
an @xmath363-geon type spacetime with a null shell can be mapped to a kruskal - type spacetime with two null shells , but these two shells must be moving in opposite directions ; our canonical transformation , on the other hand , was adapted to only one direction of the shell motion at a time . a similar issue arises if one wishes to include more than one null - dust shell .
one expects our canonical transformations to generalize readily to the case when all the shells are moving in the same direction .
shells moving in different directions would , however , seem to require new methods . several steps in our analysis relied crucially on the fact that the shell is null .
this issue appears first in the consistency of the adm equations of motion in section [ sec : metric ] . in a fixed background geometry , the equations obtained by varying the action ( [ s - ham ] ) with respect to the shell variables
must , by construction , be equivalent to the geodesic equation for the shell . in our dynamical equations ( [ metric - dyn - eoms ] ) , the pair consisting of ( [ rhat - eom ] ) and ( [ rhatmom - eom ] ) , if interpreted individually on each side of the shell , must therefore be equivalent to the null geodesic equation .
the reason why the potentially ambiguous equation ( [ rhatmom - eom ] ) turns out to be unambiguous is precisely that the junction is along a null hypersurface , and this hypersurface is geodesic in the geometries on both sides of the junction .
next , the fact that the shell history is null led us to the time as a spacetime function that is sufficiently smooth to provide an acceptable momentum conjugate to @xmath4 .
finally , in section [ sec : gauges ] , the null character of the shell history made it possible to leave the interior canonical pair @xmath248 untouched in the canonical transformation from @xmath11 to @xmath254 .
this is because the null history , when viewed from the exterior geometry , does not contain information about the interior mass , beyond the statement that @xmath247 .
these special properties of a null shell suggest that our analysis may not be immediately generalizable to timelike shells .
for example , for a dust shell with a positive rest mass , already the consistency of the adm equations of motion fails under our smoothness assumptions : the variational equations corresponding to ( [ metric - constr - eoms ] ) and ( [ metric - dyn - eoms ] ) only become consistent if the right - hand side in the counterpart of ( [ rhatmom - eom ] ) is by hand interpreted as its average over the two sides of the shell @xcite .
however , new avenues may open if one relaxes the assumption that the variations of the geometry and matter be independent .
recent progress in this direction has been made by hjek and kijowski @xcite .
the work in this paper has been purely classical .
one may , however , hope that our canonical charts on the reduced phase space will prove useful for quantizing the system . in the spatially flat gauge outside the shell
, the quantization of the shell variables with fixed interior mass was introduced as a model for hawking radiation with back - reaction in ref.@xcite , and the same approach was applied to related black holes in refs .
our results provide the tools for a similar analysis in an arbitrarily specifiable stationary gauge outside the shell .
whether this freedom in the gauge choice can be utilized to a physically interesting end remains to be seen .
one may also wish to explore quantizations based on the global canonical charts in which the dynamics is simpler but the spacetime picture more hidden .
this might shed light on the analogous question of quantizing in a dynamically simple but geometrically nontransparent canonical chart in the context of a two - dimensional dilatonic gravity theory coupled to scalar fields @xcite .
we leave these questions subject to future work .
we would like to thank tevian dray , petr hjek , jerzy kijowski , karel kucha , charlie misner , eric poisson , and stephen winters for discussions .
this work was supported in part by nsf grants phy94 - 08910 , phy94 - 21849 , and phy95 - 07740 .
in this appendix we isolate the independent information that the hamiltonian equations of motion , ( [ metric - constr - eoms ] ) and ( [ metric - dyn - eoms ] ) , contain at the shell .
it will be shown in appendices [ app : nonstatic - shell ] and [ app : static - shell ] that when this information is combined to einstein s equations away from the shell , we unambiguously recover the correct junction conditions for general relativity coupled to a null - dust shell . to begin
, we note that equations ( [ metric - constr - eoms ] ) and all save the last one of equations ( [ metric - dyn - eoms ] ) have an unambiguous distributional interpretation .
the constraint equations ( [ metric - constr - eoms ] ) contain explicit delta - functions in @xmath27 from the matter contribution and implicit delta - functions in @xmath55 and @xmath367 .
the right - hand sides of ( [ lambda - eom ] ) and ( [ r - eom ] ) contain at worst finite discontinuities , and the right - hand sides of ( [ plambda - eom ] ) and ( [ pr - eom ] ) contain at worst delta - functions ; this is consistent with the left - hand sides of ( [ lambda - eom])([pr - eom ] ) , recalling that the loci of nonsmoothness in @xmath25 , @xmath21 , @xmath53 and @xmath54 may evolve smoothly in @xmath26 .
wherever explicit or implicit delta - functions appear , they are multiplied by continuous functions of @xmath27 .
the only potentially troublesome equation is therefore ( [ rhatmom - eom ] ) : the right - hand side is a combination of spatial derivatives evaluated at the shell , but our assumptions allow these derivatives to be discontinuous .
if @xmath368 stands for any of our metric functions that may be discontinuous at the shell , we define @xmath369 \ \ .
\label{delta - def}\ ] ] the delta - contributions to @xmath370 and @xmath371 at the shell can then be written respectively as @xmath372 and @xmath373 . with this notation
, the constraint equations ( [ metric - constr - eoms ] ) at the shell read [ delta - constraints ] @xmath374 and the delta - contributions in the dynamical equations ( [ plambda - eom ] ) and ( [ pr - eom ] ) at the shell read [ delta - mom - eqs ] @xmath375 the full set of equations at the shell therefore consists of ( [ rhat - eom ] ) , ( [ rhatmom - eom ] ) , ( [ delta - constraints ] ) , and ( [ delta - mom - eqs ] ) . of these ,
all except ( [ rhatmom - eom ] ) are manifestly well defined .
two of the six equations are easily seen to be redundant .
first , inserting @xmath376 from ( [ delta - mom ] ) into ( [ delta - momlambda - eq ] ) yields an equation that is proportional to ( [ rhat - eom ] ) by the factor @xmath377 .
equation ( [ delta - momlambda - eq ] ) can therefore be dropped .
second , by continuity of the metric , we observe that @xmath378 is well defined for all @xmath26 , and so is its total time derivative , given by @xmath379^{\outwidehat } \ \ .
\label{total - der - rhat}\ ] ] the individual terms on the right - hand side of ( [ total - der - rhat ] ) are not continuous at the shell , but the left - hand side shows that the sum must be , and we obtain @xmath380 an entirely similar reasoning leads to counterparts of ( [ deltaprimedot ] ) with @xmath21 replaced by any metric function that is continuous in @xmath27 . using ( [ deltaprimedot ] ) and ( [ rhat - eom ] ) , equation ( [ plambda ] ) gives @xmath381 .
this shows that the two equations in ( [ delta - constraints ] ) are equivalent , and we can drop ( [ delta - mom ] ) . to simplify ( [ delta - momr - eq ] ) , we evaluate @xmath382 from ( [ plambda ] ) and eliminate @xmath383 and @xmath384 using ( [ deltaprimedot ] ) and its counterpart for @xmath25 . using ( [ rhat - eom ] )
, the result can be arranged to read @xmath385 \ \ , \label{delta - momr - eq - prime}\ ] ] where the vector field @xmath386 is defined by [ vee - def ] @xmath387 both at the shell and away from the shell .
at the shell , @xmath386 coincides with the shell history tangent vector @xmath59 ( [ ell - def ] ) , by virtue of the equation of motion ( [ rhat - eom ] ) . through standard manipulations , ( [ delta - momr - eq - prime ] )
can be brought to the form @xmath388 \ \ , \label{delta - momr - eq - cov}\ ] ] where @xmath389 is the spacetime covariant derivative . the information in the hamiltonian equations of motion at the shell is therefore captured by the set consisting of ( [ rhat - eom ] ) , ( [ rhatmom - eom ] ) , ( [ delta - ham ] ) , and ( [ delta - momr - eq - cov ] ) .
in appendices [ app : nonstatic - shell ] and [ app : static - shell ] we combine these four equations to the fact that away from the shell , equations ( [ metric - constr - eoms ] ) and ( [ metric - dyn - eoms ] ) are equivalent to einstein s equations and thus make the geometry locally schwarzschild .
as noted in subsection [ subsec : local - eoms ] , equation ( [ rhat - eom ] ) implies that the shell history is null .
we therefore only need to examine two qualitatively different cases , according to whether or not the shell history lies on a horizon .
the results , derived respectively in appendices [ app : nonstatic - shell ] and [ app : static - shell ] , are summarized here in the following two paragraphs : when the shell history does not lie on a horizon , the continuity of @xmath21 across the shell completely determines the geometry .
equation ( [ delta - momr - eq - cov ] ) reduces to an identity , and the combination of derivatives on the right - hand side of ( [ rhatmom - eom ] ) is continuous at the shell . equation ( [ rhatmom - eom ] ) becomes then well defined . with @xmath44 given by
( [ delta - ham ] ) , equation ( [ rhatmom - eom ] ) reduces to an identity .
when the shell history lies on a horizon , the masses on the two sides must agree .
equation ( [ delta - momr - eq - cov ] ) now implies that the soldering along the junction is affine .
the combination of derivatives on the right - hand side of ( [ rhatmom - eom ] ) is then continuous at the shell , and equation ( [ rhatmom - eom ] ) becomes well defined . with @xmath44 given by ( [ delta - ham ] ) , equation ( [ rhatmom - eom ] ) reduces to an identity .
in this appendix we verify the claims in the penultimate paragraph of appendix [ app : dist - eoms - gen ] . for concreteness , and without loss of generality
, we may assume that the shell history lies right of the horizon that the shell does not cross .
the geometry is then as in figure [ fig : nonhorizon ] for @xmath62 , and its time inverse for @xmath61 . on each side of the shell
, we introduce the coordinates , [ ef - metric ] @xmath390 where @xmath19 is the schwarzschild mass .
to avoid cluttering the notation , we suppress indices that would distinguish the coordinate patches on the two sides of the shell . wherever ambiguous quantities are encountered [ such as in equations ( [ dothatrdothatv ] ) below ] ,
the equations are understood to hold individually on each side of the shell .
the coordinates @xmath391 of section [ sec : metric ] can be embedded in the metric ( [ ef - metric ] ) as @xmath392 and @xmath393 , independently on each side of the shell .
we obtain [ g - of - vr ] @xmath394 as the surfaces of constant @xmath26 are spacelike , @xmath395 everywhere . expressing the adm variables in terms of the metric components and using ( [ g - of - vr ] ) , we find @xmath396 as in section [ sec : metric ] , the shell history is written as @xmath31 , and we write @xmath397 .
we also write , independently on each side of the shell , @xmath398 , and similarly for @xmath399 , @xmath400 , @xmath401 , @xmath402 , and so on .
we then have , independently on each side of the shell , [ dothatrdothatv ] @xmath403 with these preliminaries , we turn to the shell equations of motion . first , equation ( [ rhat - eom ] ) implies , with the help of ( [ vdotovervprime ] ) and ( [ dothatv ] ) , that @xmath404 . the shell history is therefore a hypersurface of constant @xmath334 , independently on each side .
these equations also imply that @xmath405 is unambiguous and @xmath406 equations ( [ grr - of - vr ] ) and ( [ gtr - of - vr ] ) yield , after eliminating @xmath401 and @xmath399 with the help of ( [ dothatr ] ) and ( [ rrvdotovervprime ] ) , the relation @xmath407 as the right - hand side of ( [ etavr ] ) is unambiguous , and as @xmath408 , ( [ etavr ] ) implies that @xmath402 is unambiguous .
thus , both @xmath402 and @xmath401 are unambiguous .
consider next the constraint ( [ delta - ham ] ) .
as @xmath402 and @xmath409 are unambiguous , equation ( [ grr - of - vr ] ) implies @xmath410 . using ( [ ef - metric - f ] ) , the constraint ( [ delta - ham ] ) becomes @xmath411 consider then equation ( [ rhatmom - eom ] ) .
using ( [ rrvdotovervprime ] ) and the relation @xmath412 , a straightforward calculation yields @xmath413^{\outwidehat } = { { \dot { \widehat{v ' } } } \over { \widehat{v ' } } } \ \ . \label{dothatprimevoverhatprimev}\ ] ] as the right - hand side of ( [ dothatprimevoverhatprimev ] ) is unambiguous , equations ( [ vdotovervprime ] ) and ( [ dothatprimevoverhatprimev ] ) show that the right - hand side of ( [ rhatmom - eom ] ) is unambiguous .
further , when ( [ zp - of - deltam ] ) holds , it is seen that ( [ rhatmom - eom ] ) is identically satisfied . what remains is equation ( [ delta - momr - eq - cov ] ) . in the coordinates ( [ ef - metric ] ) we have ,
again independently on the two sides of the shell , @xmath414 , and @xmath415^{\outwidehat } = - \eta \left [ { ( v^r)}^2 \ , ( \partial_r)^v,\!{}_{r } \right]^{\outwidehat } \ \ .
\label{delta - momr - cov2}\ ] ] as @xmath416 , @xmath417 is unambiguous . as @xmath418
, we see that @xmath419 is unambiguous . as the vector field denoted on each side by @xmath420
is continuous at the shell and tangent to the shell history , @xmath421^{\outwidehat}$ ] is unambiguous .
therefore , the right - hand side of ( [ delta - momr - cov2 ] ) is unambiguous , and equation ( [ delta - momr - eq - cov ] ) is identically satisfied .
in this appendix we verify the claims in the last paragraph of appendix [ app : dist - eoms - gen ] . for concreteness , and without loss of generality , we may take @xmath62 , so that the shell is moving to the left , and the geometry is as in figure [ fig : horizon ] . where @xmath19 is the common value of the schwarzschild mass .
when there is a need to distinguish the two coordinate patches , we write the coordinates as @xmath423 , with the upper ( lower ) sign referring to the patch on the right ( left ) .
the ranges of the coordinates are @xmath424 and @xmath425 , with @xmath426 , and the shell history lies at the common horizon at @xmath427 .
when the index is suppressed , equations containing ambiguous terms are understood to hold individually on each side of the shell .
consider next equation ( [ delta - momr - eq - cov ] ) . in the coordinates ( [ kruskal - metric ] ) we have , independently on the two sides of the shell , @xmath441 , and @xmath415^{\outwidehat } = \left [ { ( v^u)}^2 \ , ( \partial_r)_{u , u } \right]^{\outwidehat } \ \ .\ ] ] equation ( [ delta - momr - eq - cov ] ) therefore reads @xmath442 at the junction @xmath427 .
writing @xmath443 , @xmath444 , and @xmath445 , we obtain @xmath446 .
this means that at the junction @xmath427 we have @xmath447 where @xmath448 and @xmath449 are constants and @xmath450 .
as the kruskal coordinates are affine parameters along the horizons , this means that the soldering of the two geometries along the common horizon is affine .
consider next the constraint ( [ delta - ham ] ) . from ( [ kruskal - metric - uv ] ) we have @xmath451 .
the constraint ( [ delta - ham ] ) therefore reads @xmath452 equations ( [ grr - of - uv ] ) and ( [ gtr - of - uv ] ) yield , after eliminating @xmath453 and @xmath434 with the help of ( [ dothatu ] ) and ( [ rrvdotovervprime ] ) , the relation @xmath454 as the right - hand side of ( [ vprimedothatu ] ) is unambiguous , @xmath455 is unambiguous . the affine relation ( [ affine - us ] )
implies @xmath456 , and hence @xmath457 hence @xmath458 , and equation ( [ delta - ham - aa1 ] ) takes the form @xmath459 as @xmath460 by assumption and @xmath461 , we have @xmath462 .
this means that the right - hand - side bifurcation two - sphere occurs earlier on the history than the left - hand - side bifurcation two - sphere , as shown in figure [ fig : horizon ] .
consider finally equation ( [ rhatmom - eom ] ) . using ( [ rrvdotovervprime ] ) and proceeding as with ( [ dothatprimevoverhatprimev ] ) , we find @xmath463^{\outwidehat } = { { \dot { \widehat{v ' } } } \over { \widehat{v ' } } } \ \ . \label{dothatprimevoverhatprimev}\ ] ] by ( [ hatvplusprimeis ] ) , the right - hand side of ( [ dothatprimevoverhatprimev ] ) is unambiguous .
equations ( [ vdotovervprime ] ) and ( [ dothatprimevoverhatprimev ] ) then show that the right - hand side of ( [ rhatmom - eom ] ) is unambiguous .
when ( [ delta - ham - aa2 ] ) holds , it is seen that ( [ rhatmom - eom ] ) is identically satisfied .
r. laflamme , in _ origin and early history of the universe : proceedings of the 26th lige international astrophysical colloquium ( 1986 ) _ , edited by j. demaret ( universit de lige , institut dastrophysique , 1987 ) ; r. laflamme , ph.d .
thesis ( university of cambridge,1988 ) ; b. f. whiting and j. w. york , phys .
* 61 * , 1336 ( 1988 ) ; h. w. braden , j. d. brown , b. f. whiting , and j. w. york , phys .
d * 42 * , 3376 ( 1990 ) ; j. j. halliwell and j. louko , phys .
d * 42 * , 3997 ( 1990 ) ; g. hayward and j. louko , phys . rev .
d * 42 * , 4032 ( 1990 ) ; j. louko and b. f. whiting , class .
quantum grav .
* 9 * , 457 ( 1992 ) ; j. melmed and b. f. whiting , phys .
d * 49 * , 907 ( 1994 ) ; s. carlip and c. teitelboim , class .
quantum grav .
* 12 * , 1699 ( 1995 ) ( gr - qc/9312002 ) ; s. carlip and c. teitelboim , phys . rev .
d * 51 * , 622 ( 1995 ) ( gr - qc/9405070 ) ; g. oliveira - neto , phys .
d * 53 * , 1977 ( 1996 ) . | we consider the hamiltonian dynamics of spherically symmetric einstein gravity with a thin null - dust shell , under boundary conditions that fix the evolution of the spatial hypersurfaces at the two asymptotically flat infinities of a kruskal - like manifold .
the constraints are eliminated via a kucha - type canonical transformation and hamiltonian reduction .
the reduced phase space @xmath0 consists of two disconnected copies of @xmath1 , each associated with one direction of the shell motion .
the right - moving and left - moving test shell limits can be attached to the respective components of @xmath0 as smooth boundaries with topology @xmath2 .
choosing the right - hand - side and left - hand - side masses as configuration variables provides a global canonical chart on each component of @xmath0 , and renders the hamiltonian simple , but encodes the shell dynamics in the momenta in a convoluted way .
choosing the shell curvature radius and the interior " mass as configuration variables renders the shell dynamics transparent in an arbitrarily specifiable stationary gauge exterior " to the shell , but the resulting local canonical charts do not cover the three - dimensional subset of @xmath0 that corresponds to a horizon - straddling shell . when the evolution at the infinities is freed by introducing parametrization clocks , we find on the unreduced phase space a global canonical chart that completely decouples the physical degrees of freedom from the pure gauge degrees of freedom . replacing one infinity by a flat interior leads to analogous results , but with the reduced phase space @xmath3 .
the utility of the results for quantization is discussed .
= 10000 |
classical novae are a thermonuclear runaway event on a white dwarf ( wd ) in a binary system , in which the wd accretes hydrogen - rich matter from the companion star .
when the accreted matter reaches a critical value , hydrogen at the bottom of the wd envelope ignites to trigger a shell flash ( denoted by time @xmath15 ) . just after a nova outburst
, the envelope on the wd rapidly expands to a giant size and optically thick winds blow .
the wd envelope settles in a steady - state .
the optical magnitude attains the peak at the maximum expansion of the photosphere ( time @xmath16 ) .
then the photosphere gradually shrinks whereas the total luminosity is almost constant during the outburst .
thus , the photospheric temperature @xmath17 increases with time .
the main emitting wavelength region moves from optical to ultraviolet and finally to supersoft x - ray , corresponding to from @xmath18 k through @xmath19 k. the optically thick winds blow continuously from the very beginning of the outburst until the photospheric temperature increases to @xmath20 ( k)@xmath21 .
just when the optically thick wind stops ( time @xmath22 ) , the temperature quickly rises and the supersoft x - ray phase starts ( e.g. , * ? ? ?
thus , classical novae become a transient supersoft x - ray source in a later phase of the outburst , but their x - ray detection is rather rare mainly because of sparse observing time of x - ray satellites ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if there is a formula to predict a supersoft x - ray phase of novae , we can plan more efficient observation .
the aim of this paper is to propose such a formula .
novae show a rich variety of light curves ; the timescales of decline , shapes , and total durations of outbursts are very different from nova to nova .
therefore , it is difficult to predict a later supersoft x - ray phase only from the early optical light curve .
recently we have proposed the universal decline law of classical novae @xcite , i.e. , there is a basic light curve which can basically follow ( or overlap with ) individual nova light curves by squeezing or stretching it in the direction of time .
we call this `` a universal decline law . ''
the universal shape of light curves is almost independent of the wd mass and of the chemical composition of its hydrogen - rich envelope .
this implies that each nova light curve can be uniquely specified by one parameter ( timescaling factor squeezing / stretching factor against a universal shape of nova light curves ) .
we have reproduced a number of light curves of novae from early phase to supersoft x - ray phase ( turn - on / turnoff times ) @xcite . based on these experiences we are now able to construct empirical formulae that tell us the emergence / decay times of a supersoft x - ray phase of individual novae from observational information of the early phase .
we hope that these formulae are useful for future x - ray observations of classical novae .
the next section ( section [ modeling_of_nova ] ) describes our numerical method for obtaining x - ray , ultraviolet ( uv ) , optical , and infrared ( ir ) light curves . in section [ properties_of_light_curve ] ,
we show homologous shapes of our model light curves . using characteristic properties of free - free emission model light curves ,
we derive the absolute magnitude of each optical light curve , which provides a unique standard candle of novae . we further propose another method for obtaining the nova absolute brightness by fitting a nova light curve with the other the brightness of which
after various nova timescales and relations between these timescales are presented , we propose our prediction formulae of a supersoft x - ray phase in section [ various_timescales ] .
then , we present our theoretical understanding on the so - called maximum magnitude vs. rate of decline ( mmrd ) relation of classical novae in section [ mmrd_relation ] .
the light - curve analyses of ten classical novae are presented in section [ light_curve_of_ten_novae ] .
the x - ray turn - on / turnoff times of these ten novae , i.e. , v598 pup , v382 vel , v4743 sgr , v1281 sco , v597 pup , v1494 aql , v2467 cyg , v5116 sgr , v574 pup , and v458 vul , are consistent with the above relations ( formulae ) .
conclusions follow in sections [ conclusions ] .
the decay phase of novae can be followed by a sequence of steady state solutions ( e.g. , * ? ? ?
using the same method and numerical techniques as in @xcite , we have followed evolutions of novae by connecting steady state solutions along the decreasing envelope mass sequence .
the mass of the hydrogen - rich envelope is decreasing due to wind mass - loss and nuclear burning .
we solve a set of equations , consisting of the continuity , equation of motion , radiative diffusion , and conservation of energy , from the bottom of the hydrogen - rich envelope through the photosphere assuming spherical symmetry .
winds are accelerated deep inside the photosphere so that they are called `` optically thick winds . ''
we have presented a unified model for ir , optical , uv , and supersoft x - ray light curves of the decay phase of classical novae @xcite .
the optical and ir luminosity are reproduced by free - free emission from optically thin plasma ( e.g. , * ? ? ?
* ; * ? ? ?
this free - free emission flux at the frequency @xmath23 is estimated from equation ( 9 ) of @xcite , i.e. , @xmath24 where @xmath25 is the wind mass - loss rate , @xmath26 the wind velocity at the photosphere , and @xmath27 the photospheric radius , all of which are taken from our wind solutions .
it should be noted that the shapes of free - free emission light curves are independent of the frequency ( wavelength ) .
after the optically thick wind stops , the total mass of the ejecta remains constant in time .
the flux from such homologously expanding ejecta is estimated from equation ( 19 ) of @xcite , i.e. , @xmath28 where @xmath29 is the time after the outburst . assuming a blackbody spectrum with the photospheric temperature of @xmath17 , we have calculated the uv 1455 light curve for a narrow energy band of 14451465 ( see , e.g. , * ? ? ?
* ; * ? ? ?
* for more details ) and have also estimated the supersoft x - ray flux @xcite throughout the present paper for the energy range of 0.20.6 kev ( e.g. , * ? ? ?
even if we adopt a different energy range of 0.22.0 kev , shapes of supersoft x - ray light curves hardly change because the blackbody photosphere emits photons with energy of @xmath30 kev for @xmath31 .
the light curves of our optically thick wind model are parameterized by the wd mass ( @xmath32 ) , chemical composition of the wd envelope ( @xmath33 , @xmath34 , @xmath35 , @xmath36 , and @xmath37 ) , and envelope mass ( @xmath38 ) at the time of the outburst ( @xmath15 ) . here
@xmath33 is the hydrogen , @xmath34 the helium , @xmath35 the carbon - nitrogen - oxygen , @xmath36 the neon , and @xmath37 the metallicity content of the envelope .
we adopt @xmath39 , which also includes carbon , nitrogen , oxygen , and neon with solar composition ratios .
table [ chemical_composition ] shows nine sets of chemical compositions assumed in our models .
the details of our model light curves and their parameter dependence are described in @xcite .
llllll co nova 1 & 0.35 & 0.13 & 0.50 & 0.0 & 0.02 + co nova 2 & 0.35 & 0.33 & 0.30 & 0.0 & 0.02 + co nova 3 & 0.45 & 0.18 & 0.35 & 0.0 & 0.02 + co nova 4 & 0.55 & 0.23 & 0.20 & 0.0 & 0.02 + co nova 5 & 0.65 & 0.27 & 0.06 & 0.0 & 0.02 + ne nova 1 & 0.35 & 0.33 & 0.20 & 0.10 & 0.02 + ne nova 2 & 0.55 & 0.30 & 0.10 & 0.03 & 0.02 + ne nova 3 & 0.65 & 0.27 & 0.03 & 0.03 & 0.02 + solar & 0.70 & 0.28 & 0.0 & 0.0 & 0.02 lllllllllllllll 1.50 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 & 0.000 + 1.75 & 1.319 & 1.421 & 1.286 & 1.319 & 1.231 & 1.187 & 1.123 & 1.120 & 1.032 & 0.970 & 0.922 & 0.870 & 0.772 & 0.717 + 2.00 & 2.921 & 2.835 & 2.615 & 2.705 & 2.442 & 2.367 & 2.243 & 2.210 & 2.027 & 1.910 & 1.811 & 1.660 & 1.528 & 1.428 + 2.25 & 4.634 & 4.393 & 4.162 & 4.102 & 3.737 & 3.548 & 3.343 & 3.220 & 2.997 & 2.850 & 2.660 & 2.440 & 2.247 & 2.140 + 2.50 & 6.346 & 6.132 & 5.758 & 5.482 & 5.069 & 4.750 & 4.453 & 4.220 & 3.917 & 3.780 & 3.485 & 3.200 & 2.940 & 2.820 + 2.75 & 8.187 & 7.872 & 7.358 & 6.842 & 6.398 & 5.960 & 5.543 & 5.220 & 4.837 & 4.700 & 4.280 & 3.940 & 3.622 & 3.470 + 3.00 & 10.46 & 9.822 & 9.138 & 8.272 & 7.738 & 7.187 & 6.623 & 6.220 & 5.747 & 5.620 & 5.057 & 4.650 & 4.276 & 4.100 + 3.25 & 12.72 & 12.05 & 11.06 & 9.992 & 9.088 & 8.409 & 7.713 & 7.170 & 6.627 & 6.480 & 5.797 & 5.340 & 4.926 & 4.710 + 3.50 & 16.25 & 14.32 & 12.99 & 11.71 & 10.55 & 9.659 & 8.793 & 8.130 & 7.517 & 7.260 & 6.527 & 6.010 & 5.548 & 5.310 + 3.75 & 21.22 & 17.04 & 15.39 & 13.50 & 12.15 & 10.98 & 9.913 & 9.100 & 8.407 & 8.060 & 7.257 & 6.670 & 6.169 & 5.890 + 4.00 & 27.67 & 19.96 & 17.88 & 15.66 & 13.80 & 12.32 & 11.05 & 10.07 & 9.257 & 8.850 & 7.967 & 7.320 & 6.774 & 6.470 + 4.25 & 34.30 & 23.75 & 20.45 & 17.95 & 15.47 & 13.69 & 12.21 & 11.06 & 10.13 & 9.650 & 8.677 & 7.970 & 7.369 & 7.040 + 4.50 & 41.31 & 28.68 & 23.06 & 20.29 & 17.18 & 15.20 & 13.42 & 12.08 & 11.02 & 10.44 & 9.387 & 8.620 & 7.969 & 7.610 + 4.75 & 48.32 & 34.46 & 27.56 & 22.96 & 19.69 & 16.79 & 14.77 & 13.13 & 11.97 & 11.24 & 10.11 & 9.280 & 8.579 & 8.160 + 5.00 & 55.67 & 42.32 & 33.29 & 26.67 & 22.73 & 18.44 & 16.29 & 14.27 & 12.95 & 12.09 & 10.87 & 9.950 & 9.209 & 8.740 + 5.25 & 65.40 & 50.31 & 39.71 & 32.68 & 25.92 & 21.12 & 17.85 & 15.78 & 14.08 & 12.99 & 11.67 & 10.66 & 9.899 & 9.340 + 5.50 & 75.12 & 58.89 & 46.20 & 38.40 & 29.35 & 24.37 & 20.06 & 17.41 & 15.58 & 13.98 & 12.56 & 11.44 & 10.73 & 10.11 + 5.75 & 87.91 & 68.64 & 52.69 & 43.44 & 33.25 & 27.64 & 23.20 & 19.35 & 17.32 & 15.21 & 13.65 & 12.51 & 11.60 & 10.99 + 6.00 & 101.8 & 78.86 & 60.29 & 48.43 & 37.47 & 30.94 & 25.89 & 21.74 & 19.04 & 16.97 & 15.15 & 13.97 & 12.77 & 11.93 + 6.25 & 117.5 & 90.97 & 68.44 & 54.61 & 41.95 & 33.84 & 28.14 & 23.89 & 20.85 & 18.59 & 16.54 & 14.98 & 13.83 & 12.90 + 6.50 & 134.6 & 103.7 & 76.86 & 61.70 & 46.86 & 37.24 & 30.55 & 25.75 & 22.50 & 19.92 & 17.56 & 15.92 & 14.53 & 13.59 + 6.75 & 152.9 & 117.3 & 87.51 & 69.39 & 52.11 & 41.14 & 33.37 & 27.75 & 24.15 & 21.28 & 18.61 & 16.71 & 15.21 & 14.14 + 7.00 & 171.3 & 131.5 & 98.25 & 77.79 & 57.84 & 45.35 & 36.66 & 30.11 & 25.92 & 22.68 & 19.72 & 17.57 & 15.94 & 14.71 + 7.25 & 190.6 & 146.1 & 109.6 & 86.81 & 64.22 & 49.91 & 40.23 & 32.84 & 27.97 & 24.18 & 20.91 & 18.58 & 16.74 & 15.34 + 7.50 & 211.6 & 161.1 & 121.3 & 96.18 & 71.09 & 54.93 & 44.08 & 35.79 & 30.29 & 25.87 & 22.23 & 19.67 & 17.58 & 16.03 + 7.75 & 234.6 & 177.5 & 133.6 & 105.6 & 78.50 & 60.32 & 48.19 & 38.77 & 32.77 & 27.76 & 23.70 & 20.82 & 18.49 & 16.75 + 8.00 & 260.1 & 196.0 & 146.7 & 115.7 & 86.05 & 66.19 & 52.32 & 41.90 & 35.42 & 29.80 & 25.23 & 22.03 & 19.46 & 17.49 + 8.25 & 288.9 & 217.2 & 160.5 & 127.1 & 94.01 & 72.60 & 56.72 & 45.29 & 38.05 & 31.91 & 26.86 & 23.28 & 20.44 & 18.30 + 8.50 & 322.8 & 241.2 & 177.7 & 139.7 & 102.9 & 79.60 & 61.60 & 48.95 & 40.89 & 34.14 & 28.56 & 24.53 & 21.47 & 19.12 + 8.75 & 363.7 & 268.6 & 197.2 & 153.9 & 113.1 & 87.22 & 66.89 & 53.02 & 44.02 & 36.54 & 30.38 & 25.89 & 22.58 & 19.97 + 9.00 & 411.1 & 301.4 & 219.4 & 171.6 & 124.3 & 95.52 & 73.07 & 57.43 & 47.44 & 39.14 & 32.31 & 27.46 & 23.81 & 20.94 + 9.25 & 457.6 & 339.0 & 246.2 & 191.7 & 138.9 & 104.5 & 80.58 & 62.78 & 51.34 & 42.10 & 34.43 & 29.22 & 25.19 & 22.04 + 9.50 & 508.2 & 377.0 & 276.1 & 214.1 & 155.7 & 117.0 & 89.01 & 69.27 & 56.25 & 45.37 & 37.10 & 31.32 & 26.89 & 23.32 + 9.75 & 547.2 & 412.0 & 307.1 & 239.8 & 173.7 & 131.4 & 100.0 & 76.87 & 62.24 & 49.59 & 40.33 & 33.85 & 28.93 & 24.79 + 10.00 & 589.5 & 446.5 & 335.6 & 267.7 & 193.2 & 146.7 & 112.5 & 86.46 & 69.49 & 54.71 & 44.21 & 37.08 & 31.56 & 26.67 + 10.25 & 635.0 & 479.3 & 362.2 & 292.3 & 215.6 & 163.5 & 124.7 & 97.02 & 78.11 & 60.81 & 48.97 & 40.76 & 34.69 & 29.01 + 10.50 & 673.6 & 512.3 & 386.2 & 313.5 & 234.6 & 180.4 & 138.3 & 107.1 & 86.93 & 68.04 & 54.70 & 45.39 & 38.36 & 31.74 + 10.75 & 715.8 & 541.6 & 412.6 & 335.6 & 253.3 & 195.2 & 153.4 & 118.5 & 95.91 & 75.64 & 60.98 & 50.50 & 42.51 & 34.85 + 11.00 & 762.2 & 571.6 & 440.2 & 355.9 & 269.9 & 209.3 & 165.7 & 130.3 & 105.7 & 83.14 & 66.78 & 55.83 & 46.53 & 38.24 + 11.25 & 813.6 & 603.7 & 464.6 & 377.3 & 284.6 & 223.5 & 177.7 & 140.4 & 114.3 & 91.54 & 73.13 & 60.75 & 50.34 & 41.48 + 11.50 & 869.2 & 637.0 & 489.4 & 396.0 & 300.8 & 237.0 & 187.3 & 149.9 & 121.1 & 99.24 & 79.95 & 65.55 & 54.40 & 44.25 + 11.75 & 920.8 & 673.7 & 516.6 & 415.2 & 317.0 & 249.3 & 197.7 & 158.9 & 128.5 & 105.0 & 84.82 & 70.35 & 58.73 & 47.19 + 12.00 & 975.4 & 713.8 & 546.4 & 436.2 & 332.6 & 261.8 & 209.0 & 167.0 & 136.5 & 111.3 & 89.90 & 74.95 & 62.18 & 50.27 + 12.25 & 1034 .
& 756.3 & 578.9 & 459.2 & 349.7 & 274.1 & 219.3 & 175.8 & 145.2 & 118.1 & 95.36 & 79.25 & 65.64 & 53.55 + 12.50 & 1095 . &
801.4 & 613.5 & 484.3 & 368.4 & 287.4 & 229.3 & 184.0 & 154.3 & 125.4 & 101.1 & 83.80 & 69.32 & 56.94 + 12.75 & 1160 . & 849.1 & 650.2 & 513.2 & 388.7 & 302.1 & 240.2 & 192.7 & 161.7 & 133.1 & 107.3 & 88.40 & 73.16 & 60.21 + 13.00 & 1229 . & 899.8 & 689.0 & 544.1 & 411.5 & 317.9 & 252.1 & 202.1 & 169.7 & 139.7 & 113.6 & 93.20 & 77.25 & 63.71 + 13.25 & 1302 . & 953.3 & 730.1 & 576.7 & 436.1 & 336.6 & 265.4 & 212.3 & 178.3 & 146.4 & 120.7 & 98.30 & 81.99 & 67.40 + 13.50 & 1379 . & 1011 . & 773.7 & 611.3 & 462.2 & 356.7 & 281.6 & 224.3 & 187.5 & 153.8 & 128.1 & 103.8 & 87.01 & 71.28 + 13.75 & 1461 . & 1071 .
& 819.8 & 647.9 & 489.8 & 377.9 & 298.9 & 238.4 & 199.1 & 161.6 & 136.0 & 109.6 & 92.33 & 75.31 + 14.00 & 1547 . &
1134 . & 868.7 & 686.7 & 519.0 & 400.5 & 317.1 & 253.2 & 211.4 & 173.1 & 144.4 & 116.8 & 97.94 & 80.17 + 14.25 & 1639 . & 1202 . & 920.5 & 727.8 & 550.0 & 424.3 & 336.5 & 269.0 & 224.4 & 185.5 & 153.2 & 124.8 & 103.9 & 85.41 + 14.50 & 1736 . & 1273 . & 975.3 & 771.3 & 582.9 & 449.6 & 357.0 & 285.7 & 238.2 & 198.6 & 162.6 & 133.3 & 110.2 & 90.96 + 14.75 & 1839 . & 1349 .
& 1034 . & 817.4 & 617.7 & 476.4 & 378.6 & 303.4 & 252.9 & 212.4 & 172.6 & 142.3 & 116.9 & 96.87 + 15.00 & 1948 . & 1429 . & 1095 .
& 866.2 & 654.5 & 504.8 & 401.6 & 322.1 & 268.3 & 227.1 & 183.1 & 151.8 & 124.0 & 103.1 + x - ray & 3850 & 2700 & 2200 & 1700 & 1150 & 870 & 620 & 440 & 326 & 240 & 162 & 109 & 74.1 & 48.7 + @xmath40 & 17.6 & 17.3 & 17.0 & 16.8 & 16.6 & 16.3 & 16.2 & 16.1 & 15.9 & 15.9 & 15.7 & 15.7 & 15.8 & 15.7 + @xmath41 & 0.81 & 0.72 & 0.62 & 0.52 & 0.42 & 0.31 & 0.21 & 0.10 & 0.01 & @xmath42 & @xmath43 & @xmath44 & @xmath45 & @xmath46 + @xmath47 & 5.3 & 4.8 & 4.3 & 3.8 & 3.3 & 2.8 & 2.4 & 2.0 & 1.6 & 1.4 & 1.0 & 0.8 & 0.6 & 0.2 + llllllllllllll 3.000 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 & 0.0 + 3.250 & 2.830 & 2.270 & 1.610 & 1.292 & 1.120 & 0.960 & 0.870 & 0.758 & 0.667 & 0.608 & 0.565 & 0.515 & 0.471 + 3.500 & 6.860 & 4.590 & 3.230 & 2.742 & 2.270 & 1.940 & 1.740 & 1.499 & 1.352 & 1.215 & 1.117 & 1.026 & 0.939 + 3.750 & 10.98 & 6.970 & 5.070 & 4.292 & 3.450 & 2.960 & 2.630 & 2.259 & 2.005 & 1.828 & 1.669 & 1.535 & 1.408 + 4.000 & 15.27 & 9.400 & 7.630 & 5.882 & 4.770 & 4.020 & 3.560 & 3.035 & 2.683 & 2.429 & 2.227 & 2.039 & 1.872 + 4.250 & 19.66 & 12.35 & 10.34 & 7.572 & 6.260 & 5.150 & 4.500 & 3.874 & 3.372 & 3.052 & 2.784 & 2.543 & 2.336 + 4.500 & 24.59 & 16.54 & 13.12 & 9.872 & 7.810 & 6.460 & 5.480 & 4.736 & 4.154 & 3.716 & 3.429 & 3.114 & 2.863 + 4.750 & 30.87 & 20.97 & 16.18 & 12.51 & 9.470 & 8.050 & 6.870 & 5.736 & 4.981 & 4.466 & 4.097 & 3.717 & 3.480 + 5.000 & 38.11 & 25.58 & 19.56 & 15.25 & 11.51 & 9.680 & 8.330 & 6.980 & 5.865 & 5.276 & 4.817 & 4.346 & 4.068 + 5.250 & 45.78 & 31.11 & 23.10 & 17.98 & 13.69 & 11.34 & 9.740 & 8.270 & 6.945 & 6.126 & 5.557 & 4.996 & 4.600 + 5.500 & 54.00 & 37.61 & 26.89 & 20.83 & 15.97 & 13.10 & 11.05 & 9.340 & 7.935 & 7.046 & 6.297 & 5.616 & 5.119 + 5.750 & 65.14 & 44.69 & 32.20 & 23.82 & 18.41 & 15.02 & 12.46 & 10.38 & 8.855 & 7.846 & 6.997 & 6.186 & 5.588 + 6.000 & 78.08 & 52.44 & 38.16 & 27.74 & 21.03 & 17.16 & 14.13 & 11.57 & 9.745 & 8.566 & 7.647 & 6.756 & 6.043 + 6.250 & 92.06 & 62.15 & 44.65 & 32.40 & 24.08 & 19.48 & 16.08 & 13.08 & 10.83 & 9.416 & 8.327 & 7.326 & 6.502 + 6.500 & 106.9 & 72.79 & 51.93 & 37.53 & 27.75 & 22.27 & 18.22 & 14.77 & 12.09 & 10.37 & 9.087 & 7.886 & 6.973 + 6.750 & 121.6 & 84.36 & 59.93 & 43.19 & 31.76 & 25.48 & 20.73 & 16.64 & 13.50 & 11.46 & 9.937 & 8.546 & 7.467 + 7.000 & 137.5 & 95.87 & 68.63 & 49.46 & 36.14 & 28.98 & 23.51 & 18.75 & 15.04 & 12.65 & 10.86 & 9.206 & 7.997 + 7.250 & 153.6 & 107.6 & 77.53 & 56.30 & 40.95 & 32.66 & 26.49 & 21.09 & 16.78 & 13.95 & 11.87 & 9.956 & 8.537 + 7.500 & 171.0 & 120.3 & 86.47 & 63.47 & 46.23 & 36.67 & 29.55 & 23.51 & 18.58 & 15.37 & 12.98 & 10.72 & 9.087 + 7.750 & 190.3 & 134.1 & 96.17 & 70.30 & 51.94 & 41.04 & 32.87 & 25.93 & 20.38 & 16.77 & 14.09 & 11.53 & 9.627 + 8.000 & 214.2 & 149.1 & 107.7 & 77.72 & 57.59 & 45.66 & 36.49 & 28.49 & 22.32 & 18.22 & 15.23 & 12.35 & 10.19 + 8.250 & 241.0 & 167.0 & 120.4 & 87.02 & 63.69 & 50.42 & 40.36 & 31.26 & 24.43 & 19.81 & 16.44 & 13.18 & 10.77 + 8.500 & 270.1 & 187.9 & 134.9 & 97.51 & 71.21 & 55.63 & 44.52 & 34.40 & 26.85 & 21.55 & 17.74 & 14.12 & 11.41 + 8.750 & 301.9 & 211.4 & 151.8 & 109.3 & 79.80 & 62.42 & 49.45 & 38.30 & 29.64 & 23.69 & 19.26 & 15.25 & 12.18 + 9.000 & 338.4 & 235.7 & 170.6 & 122.9 & 89.49 & 69.91 & 55.54 & 42.96 & 32.86 & 26.07 & 21.07 & 16.60 & 13.15 + 9.250 & 376.5 & 263.7 & 191.0 & 138.2 & 100.7 & 78.89 & 62.33 & 48.14 & 36.75 & 29.04 & 23.32 & 18.17 & 14.32 + 9.500 & 409.3 & 295.3 & 214.2 & 155.4 & 113.3 & 89.13 & 70.38 & 54.26 & 41.14 & 32.40 & 25.95 & 20.02 & 15.61 + 9.750 & 447.0 & 320.0 & 240.5 & 175.0 & 127.4 & 100.4 & 79.33 & 61.20 & 46.37 & 36.42 & 28.95 & 22.15 & 17.15 + 10.00 & 490.2 & 347.6 & 259.3 & 196.8 & 143.3 & 113.2 & 89.49 & 69.13 & 52.20 & 40.94 & 32.36 & 24.74 & 19.00 + 10.25 & 539.6 & 379.0 & 280.5 & 211.7 & 161.0 & 127.6 & 100.9 & 78.49 & 59.10 & 46.07 & 36.50 & 27.61 & 21.04 + 10.50 & 577.7 & 414.4 & 304.3 & 228.4 & 173.3 & 140.5 & 113.9 & 87.86 & 66.88 & 51.98 & 41.46 & 30.91 & 23.52 + 10.75 & 612.7 & 448.9 & 331.2 & 247.1 & 186.9 & 151.2 & 123.1 & 97.18 & 74.73 & 58.50 & 46.47 & 34.56 & 26.20 + 11.00 & 651.7 & 474.6 & 358.2 & 267.9 & 202.1 & 162.9 & 132.6 & 106.4 & 82.36 & 64.22 & 51.36 & 38.45 & 28.83 + 11.25 & 694.6 & 502.9 & 378.2 & 289.6 & 218.9 & 175.9 & 143.1 & 115.0 & 89.13 & 69.31 & 56.29 & 41.87 & 31.37 + 11.50 & 741.9 & 534.3 & 400.3 & 305.4 & 236.6 & 190.1 & 154.6 & 122.7 & 95.41 & 74.79 & 60.80 & 45.47 & 33.74 + 11.75 & 787.7 & 568.8 & 424.7 & 322.8 & 249.5 & 203.3 & 167.1 & 131.0 & 102.1 & 80.68 & 65.21 & 49.08 & 36.17 + 12.00 & 835.3 & 605.2 & 451.6 & 342.0 & 263.7 & 214.6 & 177.0 & 139.3 & 109.3 & 86.96 & 69.69 & 52.43 & 38.69 + 12.25 & 885.6 & 641.6 & 479.8 & 363.0 & 279.2 & 227.0 & 186.9 & 147.6 & 116.1 & 93.09 & 74.47 & 55.96 & 41.36 + 12.50 & 939.0 & 680.3 & 509.2 & 385.4 & 296.3 & 240.7 & 197.7 & 156.7 & 123.4 & 98.90 & 79.37 & 59.84 & 44.29 + 12.75 & 995.4 & 721.2 & 540.4 & 408.7 & 314.7 & 255.4 & 209.5 & 166.6 & 131.2 & 105.2 & 84.67 & 63.97 & 47.35 + 13.00 & 1055 .
& 764.5 & 573.4 & 433.5 & 334.1 & 271.2 & 222.4 & 177.1 & 139.6 & 112.1 & 90.29 & 68.34 & 50.60 + 13.25 & 1118 .
& 810.4 & 608.4 & 459.7 & 354.7 & 287.9 & 236.2 & 187.8 & 148.3 & 119.2 & 96.26 & 72.90 & 53.86 + 13.50 & 1185 .
& 859.0 & 645.4 & 487.5 & 376.5 & 305.7 & 250.9 & 199.2 & 157.4 & 126.6 & 102.3 & 77.56 & 57.21 + 13.75 & 1257 .
& 910.6 & 684.6 & 517.0 & 399.6 & 324.4 & 266.4 & 211.2 & 167.0 & 134.5 & 108.8 & 82.50 & 60.77 + 14.00 & 1332 .
& 965.1 & 726.2 & 548.2 & 424.0 & 344.4 & 282.9 & 223.9 & 177.2 & 142.9 & 115.6 & 87.73 & 64.53 + 14.25 & 1412 . & 1023 .
& 770.2 & 581.2 & 450.0 & 365.4 & 300.4 & 237.4 & 188.0 & 151.7 & 122.9 & 93.27 & 68.52 + 14.50 & 1496 . & 1084 .
& 816.8 & 616.2 & 477.4 & 387.7 & 318.8 & 251.7 & 199.4 & 161.1 & 130.5 & 99.16 & 72.75 + 14.75 & 1586 . & 1149 .
& 866.2 & 653.2 & 506.5 & 411.4 & 338.4 & 266.8 & 211.5 & 171.1 & 138.6 & 105.4 & 77.22 + 15.00 & 1681 . & 1218 .
& 918.5 & 692.5 & 537.2 & 436.4 & 359.1 & 282.9 & 224.3 & 181.6 & 147.2 & 112.0 & 81.96 + x - ray & 4300 & 3100 & 2240 & 1620 & 1170 & 864 & 640 & 440 & 300 & 201 & 135 & 79.3 & 42.2 + @xmath40 & 15.2 & 14.9 & 14.8 & 14.5 & 14.3 & 14.2 & 14.0 & 14.0 & 13.8 & 13.8 & 13.8 & 13.8 & 13.8 + @xmath41 & 0.72 & 0.60 & 0.48 & 0.38 & 0.27 & 0.18 & 0.09 & @xmath48 & @xmath49 & @xmath50 & @xmath51 & @xmath52 & @xmath53 + @xmath47 & 4.7 & 4.1 & 3.7 & 3.1 & 2.7 & 2.3 & 2.0 & 1.6 & 1.2 & 0.9 & 0.7 & 0.4 & 0.0 + the decay timescale of a light curve depends mainly on the wd mass and weakly on the chemical composition , whereas the maximum brightness of a light curve depends strongly on the initial envelope mass , @xmath38 . in this section
, we present various features , especially the absolute magnitudes , of nova light curves based on our universal decline law .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] shows our free - free emission model light curves calculated from equation ( [ wind - free - free - emission ] ) , i.e. , @xmath54^{\{m_{\rm wd}\}}_{(t ) } + ~ g^{\{m_{\rm wd}\ } } , \label{template - wind - free - free - emission}\ ] ] for various wd masses , from @xmath55 to @xmath56 by step of @xmath57 for a chemical composition of `` co nova 2 '' ( these numerical data are tabulated in table [ light_curves_of_novae_co ] ) .
the subscript @xmath58 explicitly expresses that this is a function of time while the superscript @xmath59 indicates a model parameter of wd mass .
an optically thick wind phase continues for a long time , from a very early phase to a late phase of a nova outburst . in this figure
we plot free - free emission model light curves calculated by equation ( [ template - wind - free - free - emission ] ) from a very early phase until the end of an optically thick wind phase .
the 15th mag point of each light curve corresponds to the end of the wind phase . in other words
, we define a constant of @xmath60 in equation ( [ template - wind - free - free - emission ] ) as that the last ( lowest ) point of each light curve ( the end of an optically thick wind phase ) is 15th mag . as a result , the value of @xmath60 is different for different @xmath32 . in order to make table [ light_curves_of_novae_co ] more compact , we have defined @xmath60 in this way . these light curves demonstrate that the decay timescales of novae are sensitive to the wd mass
. therefore we can determine the wd mass of a nova from light curve fitting .
the @xmath61 ( blue thick solid line ) or @xmath62 wd may be a typical case of fast novae and the light curve of @xmath63 wd ( red thick solid line ) may correspond to a slow nova .
these free - free emission model light curves are applicable to optical and near infrared ( ir ) wavelength regions ( e.g. , * ? ? ?
we added observational @xmath64 magnitudes of v1668 cyg ( nova cygni 1978 ) to figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] , which nicely reproduce the light curve of @xmath61 wd . in this case , we shift the model light curve of @xmath61 wd down by 1.2 mag in order to fit it with the observation ( see the right axis of figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] ) .
the intermediate width @xmath64 band is an emission line - free optical band , so it is ideal to fit our model light curve with @xmath64 magnitudes . in this way
, we can shift the template light curves to fit them with observation for individual novae .
in other words , one can shift the model light curve `` up and down '' against the observational data of an individual nova . in the case of v1668 cyg
, we obtain @xmath65 from the @xmath61 wd model in table [ light_curves_of_novae_co ] .
the free - free emission model light curves in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] have strong similarity in their shapes . in our previous paper @xcite
, we showed that these model light curves are homologous and their shapes are almost overlapped with each other when they are properly squeezed or stretched along time . in this subsection
, we reformulate this homologous properties in a more sophisticated manner and obtain their brightnesses .
first , we squeeze or stretch our uv 1455 model light curves in the direction of time and also its peak flux is normalized to be fitted with the observational peak of v1668 cyg . then we obtain all the uv light curves almost overlapped to each other as shown in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
we determine a stretching factor , @xmath66 , of each uv model light curve by increasing or decreasing it until the model light curve shape matches the observation .
table [ light_curves_of_novae_co ] shows logarithmic values of @xmath41 , that is , positive ( negative ) for stretching ( squeezing ) against the v1668 cyg uv 1455 observation .
for example , we see that the @xmath61 wd model light curve is almost overlapped to the v1668 cyg observation without stretching ( @xmath67 ) and the @xmath68 wd model light curve evolves about twice slower ( @xmath69 ) than the @xmath61 wd model . in this way , all the uv 1455 model light curves
are overlapped to each other , if we squeeze / stretch the time as @xmath70 .
note that the stretching factor is slightly different from that in our previous paper ( table 9 of * ? ? ?
* ) , which was determined in the way that each free - free model light curve is overlapped with each other instead of uv 1455 model light curves .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] also shows optical data ( @xmath64 and @xmath71 magnitudes ) of v1668 cyg and fitted free - free emission model light curve of @xmath61 wd ( @xmath72 ) . the other free - free emission model light curves are squeezed / stretched with the same @xmath66 as that of uv light curves .
when we squeeze the model light curves by a factor of @xmath66 in the time direction ( @xmath73 and @xmath74 ) , the flux is also changed as @xmath75 because @xmath76 therefore , we shift the model light curves in the vertical direction as @xmath77 and obtained figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] . here
@xmath78 is the apparent @xmath71 magnitude of an @xmath32 model and @xmath79 is its time - stretched magnitude against a standard model of 0.95 @xmath80 wd ( @xmath81 v1668 cyg ) .
we see that all the light curves are almost overlapped to each other . for more details of derivation of equation ( [ simple_final_scaling_flux ] ) , see appendix a. the origins of magnitude in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] are different because the magnitude in the right axis of figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] is calibrated by the @xmath64 magnitudes whereas by the @xmath71 magnitudes in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
we define the rescaled apparent @xmath71 magnitude ( time - stretched @xmath71 magnitude ) against the @xmath71 magnitudes of v1668 cyg instead of the @xmath64 magnitudes .
the apparent @xmath71 magnitudes are written as @xmath82^{\{m_{\rm wd}\}}_{(t / f_{\rm s } ) } + k_v , \label{simple_scaling_flux}\ ] ] where @xmath83 is a constant and is determined to match @xmath79 of the 0.95 @xmath80 wd model with the @xmath71 magnitude of v1668 cyg as shown in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
therefore , @xmath83 is also related to @xmath84 as @xmath85 this simply means @xmath86 for 0.95 @xmath80 wd model with @xmath87 .
note that we here obtain 0.9 instead of 1.2 because observational @xmath78 is brighter than @xmath88 by about 0.3 mag , that is , @xmath78 and @xmath89 are overlapped to each other in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
we use the same @xmath83 for all other wd mass models ( from @xmath55 to 1.2 @xmath80 by 0.05 @xmath80 step ) in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] . all the free - free emission model light curves are almost overlapped to each other .
it should be noticed here that the flux is condensed and increased by a factor of @xmath66 when the time is stretched by a factor of @xmath90 ( or squeezed by a factor of @xmath66 ) _ only under the condition that the spectrum is not changed after time - stretching_. this condition is satisfied in free - free emission ( optical and near ir ) because the flux ( @xmath91 constant ) is independent of the frequency @xmath23 , but not satisfied in blackbody emission because the blackbody emissivity depends on the frequency @xmath23 and , after time - stretching , a blackbody spectrum changes with @xmath74 .
see appendix a for more detailed formulation of the scaling of free - free flux .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] shows three points of a , b , and c on the same light curve of @xmath61 wd model .
point b indicates the initial envelope mass of @xmath92 , corresponding to @xmath16 at the maximum brightness of v1668 cyg .
point a indicates a much brighter maximum and corresponds to a much larger envelope mass ( @xmath93 ) than that of point b. on the other hand , point c corresponds to a much smaller envelope mass ( @xmath94 ) .
the initial envelope mass depends on the mass accretion rate to the wd before ignition ( see , e.g. , figure 9 of * ? ? ?
* ; * ? ? ?
point a is corresponding to a case of much lower mass accretion rates to the wd . on the other hand , point c is corresponding to a case of much higher mass accretion rates to the wd than that of point b , for the case of v1668 cyg .
this simply suggests that novae are brighter for lower mass accretion rates . for the other wd masses ,
the trend between the maximum brightness and the initial envelope mass is the same as that of @xmath61 . for more ( less )
massive wds , the envelope mass at the same point b is smaller ( larger ) than that of @xmath61 .
we will see again these relations between the maximum brightness and the initial envelope mass in more details in section [ mmrd_relation ] .
we regard that v1668 cyg is a typical classical nova and point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] corresponds to typical brightnesses for other wd mass models .
therefore , we replot the same nova model light curves in a logarithmic time in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] . in the overlapping procedure of uv 1455 model light curves , we adopt the origin of time which is different from the origin of time in @xmath95 of table [ light_curves_of_novae_co ] .
this procedure is necessary for the horizontal axis of @xmath96 because we must start all the free - free emission light curves from the same stage like point b ( or c ) in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
here we define the origin of time as about 4 days before point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] , that is , the outburst day of v1668 cyg . in this case , the initial brightnesses of novae are almost the same ( brightness at point b ) and overlapped with each other in the @xmath96@xmath79 diagram .
we have given an apparent magnitude to each free - free emission model light curve against v1668 cyg data .
next step is to determine the absolute magnitudes of these model light curves .
this can be done by using the distance modulus of v1668 cyg .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] shows observed @xmath64 and @xmath71 magnitudes of v1668 cyg , which can be converted to the absolute magnitude from the distance modulus to v1668 cyg , i.e. , @xmath97_{\rm v1668~cyg } = 14.3 .
\label{distance_modulus_v1668_cyg}\ ] ] here , we adopt the distance of @xmath98 kpc @xcite and the absorption of @xmath99 @xcite for v1668 cyg .
since our @xmath61 wd model shows a good agreement with the @xmath64 and @xmath71 magnitudes of v1668 cyg , we determine the absolute @xmath71 magnitude of our @xmath61 wd model light curve as follows : @xmath100 where @xmath101 for @xmath61 wd model ( @xmath102 ) from equation ( [ relation_between_ff_and_mv ] ) .
then , the absolute @xmath71 magnitude at the last point of free - free template light curve ( the end point of table [ light_curves_of_novae_co ] ) corresponds to @xmath103 where the subscript of `` w '' means `` wind '' which corresponds to @xmath104 at the end point ( table [ light_curves_of_novae_co ] ) . using @xmath47 , we restore the entire absolute @xmath71 magnitudes of the @xmath61 wd model by @xmath105 and it is shown by a red solid line in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] . for the other wd mass models , from equation ( [ simple_final_scaling_flux ] )
we simply derive the following relations , i.e. , @xmath106 @xmath107 based on the v1668 cyg data .
we plot @xmath47 against various wd masses in figure [ absolute_mag_v1974cyg_v1668_cyg_scale ] . using @xmath47 , we have @xmath108 where @xmath41 , @xmath40 , and @xmath47 are tabulated in table [ light_curves_of_novae_co ] for all other wd mass models from @xmath55 to 1.2 @xmath80 by 0.05 @xmath80 step . here
prime symbol means the values for squeezed / stretched light curves .
we have restored the entire absolute @xmath71 magnitudes for @xmath109@xmath110 wds in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] .
it is interesting to examine whether or not the absolute magnitudes of our free - free emission model light curves are consistent with various empirical relations of classical novae that have ever been proposed .
very popular relations between the maximum magnitude and the rate of decline ( mmrd ) will be discussed in section [ mmrd_relation ] . here
we check the empirical formula that the absolute magnitude at 15 days after optical maximum , @xmath111 , is almost common among various novae ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this relation was first proposed by @xcite with @xmath112 , then followed by @xcite with @xmath113 , by @xcite with @xmath114 , by @xcite with @xmath115 , and by @xcite with @xmath116 .
we have obtained @xmath117 for @xmath109@xmath110 wds ( 14 light curves ) with equal weight for all wd mass models , being roughly consistent with the above empirical estimates as shown by blue crosses in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] . for the statistical point of view ,
carbon - oxygen ( co ) novae have typical masses of @xmath118@xmath119 .
if we take this rage of wd masses ( equal weight ) , the value becomes @xmath120 as indicated in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] , being close to the value proposed by @xcite .
in this subsection , we study light curves of neon ( ne ) novae and show that the essential results on the free - free emission model light curves are the same as those for co novae studied in the previous subsections .
table [ light_curves_of_novae_ne ] lists free - free model light curves in the form of @xmath95 , definition of which is the same ( see equation ( [ template - wind - free - free - emission ] ) ) but for the chemical composition of `` ne nova 2 '' in table [ chemical_composition ] .
figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] shows free - free ( optical ) , uv 1455 , and supersoft x - ray model light curves as well as observational data of v1974 cyg ( nova cygni 1992 ) .
we plot these light curves in @xmath96@xmath79 plane and show nova evolutions until more later phases . in the overlapping procedure of uv 1455 model light curves
, we adopt the origin of time which is different from the origin of time in @xmath95 of table [ light_curves_of_novae_ne ] .
this procedure is necessary for the same reason as in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] . as a result ,
the peak brightness of each light curve is almost the same in the @xmath96@xmath79 plane as shown in figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] . in this figure ,
we adopt @xmath121 for @xmath119 wd model ( @xmath72 ) as a starting point of the model light curves , corresponding to @xmath122 of @xmath119 wd model in table [ light_curves_of_novae_ne ] .
we also obtain the absolute magnitudes of free - free model light curves in the same way as the co novae mentioned in the previous subsections , but for the neon nova v1974 cyg as a comparison . adopting the distance of @xmath123 kpc @xcite and the absorption of @xmath124 @xcite for v1974 cyg
, we obtain the distance modulus of @xmath97_{\rm v1974~cyg } = 12.3 , \label{distance_modulus_v1974_cyg}\ ] ] for v1974 cyg .
then , we specify the absolute magnitude at the bottom of our free - free model light curves as @xmath125 for v1974 cyg and tabulate @xmath66 , @xmath40 , and @xmath47 in table [ light_curves_of_novae_ne ] .
we plot @xmath47 for both co and ne novae in figure [ absolute_mag_v1974cyg_v1668_cyg_scale ] .
this clearly shows that the brightness @xmath47 is smoothly increasing from @xmath126 mag to @xmath127 mag with @xmath32 from 0.55 to 1.2 @xmath80 for `` co nova 2 '' ( from 0.7 to 1.3 @xmath80 for `` ne nova 2 '' ) .
now , we can determine the absolute magnitude of free - free model light curves , @xmath128 from table [ light_curves_of_novae_ne ] , for various wd masses as shown in figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] .
we also check whether or not @xmath111 is almost common among various neon novae .
we have obtained @xmath129 for @xmath130@xmath131 wds ( 13 light curves with equal weight for all wd mass models ) , being roughly consistent with the empirical estimates ( see section [ absolute_magnitude ] ) . for the statistical point of view
, neon novae have typical masses of @xmath132@xmath133 .
if we take this rage of wd masses ( equal weight for 6 light curves ) , the value becomes @xmath134 as indicated in figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] .
if we take all the co and ne nova light curves with equal weight ( @xmath135 wd mass models ) , the average value becomes @xmath117 .
thus , the absolute magnitudes of our free - free emission model light curves are consistent with the @xmath111 empirical relations .
we have proposed new methods for obtaining the absolute magnitude of novae in the previous subsections . using these absolute magnitudes
, we can estimate the distance to novae . in this subsection
, we will check the accuracy of our new methods by comparing three nova light curves .
first of all , we explain how to select a model that reproduces well observational light curves .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] shows our method in the case of v1668 cyg .
we shift the model light curve `` back and forth '' and `` up and down '' to fit with the observational @xmath64 magnitudes . together with the uv 1455 light curve , we finally determine by eyes the @xmath61 wd as the best - reproducing one among the light curves tabulated in table [ light_curves_of_novae_co ] ( @xmath136 by @xmath57 step ) . in this figure
, we added two other light curves of different wd masses , i.e. , @xmath137 and @xmath119 wds . here
, we use different phases of the model light curves in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] to fit with the @xmath64 observation ; the @xmath138 wd model is shifted rightward ( back ) but the @xmath119 wd model is shifted leftward ( forth ) . as a result , the three model light curves of @xmath137 , @xmath139 , and @xmath119 wds give similar decline rates from maximum until day @xmath140 . in other words
, we can find similar shapes in different phases of different wd mass light curves in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_test_all ] . in this sense , it is hard to choose the best - reproducing model only from early phase optical and near ir data if neither uv 1455 nor x - ray data are available .
it should be also noted here that the ambiguity of wd mass estimate of @xmath141 introduces an error of about @xmath142 mag in the distance modulus of @xmath143 , when only early optical ( or ir ) light curves are available . in figure
[ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] , we obtain @xmath144 for @xmath145 , but @xmath146 for @xmath147 or @xmath148 for @xmath149 . now let us go to the first example of a very fast nova v1500 cyg ( nova cygni 1975 ) .
figure [ light_curve_v598_pup_v1500_cyg ] demonstrates that optical @xmath71 ( filled triangles ) , @xmath64 ( filled circles ) , near infrared @xmath150 ( open squares ) , and @xmath151 ( open triangles ) magnitudes are almost overlapped with each other from @xmath126 days until @xmath152 days after the outburst .
this wavelength - free decline is a characteristic property of free - free emission .
this impressive part is reproduced well by equation ( [ wind - free - free - emission ] ) with the models of @xmath147 ( `` co nova 2 '' ) and @xmath153 ( `` ne nova 2 '' ) .
v1500 cyg is a superbright nova , in which the peak luminosity was about 4 mag brighter than the eddington limit of a 1.0 @xmath80 wd ( e.g. , * ? ? ?
* ) , and the spectra of the superbright part ( first 5 days ) are approximated by blackbody emission ( e.g. , * ? ? ?
* ) . therefore , our free - free emission model can not be applicable to the superbright part .
after day @xmath154 , the @xmath71 light curve gradually deviates from the other light curves ( @xmath64 , @xmath150 , and @xmath151 ) because emission lines dominate the optical fluxes .
the @xmath71 magnitudes are 2 mag or more brighter than the @xmath64 , @xmath150 , and @xmath151 magnitudes in the late phase . as extensively discussed in @xcite and @xcite , @xmath71 magnitudes are often heavily contaminated by strong emission lines such as [ ] @xmath155 4959 , 5007 ( see * ? ? ? * for v598 pup spectra ) , which eventually dominate over the continuum , causing an increasing deviation from the free - free emission model . comparing the apparent magnitude of v1500 cyg and the absolute magnitude of the model light curve of @xmath119 ( table [ light_curves_of_novae_co ] )
, we can estimate the distance modulus .
figure [ light_curve_v598_pup_v1500_cyg ] shows that the @xmath119 wd model is a best - reproducing model ( magenta dash - dotted line ) among other wd mass models in table [ light_curves_of_novae_co ] for chemical composition of `` co nova 2 . ''
using @xmath156 in table [ light_curves_of_novae_co ] and @xmath157 from figure [ light_curve_v598_pup_v1500_cyg ] , we obtain the distance modulus to v1500 cyg , i.e. , @xmath158 .
if we adopt another chemical composition of `` ne nova 2 , '' we obtain a best - reproducing model of @xmath133 wd ( black solid line ) . then we obtain the distance modulus of @xmath159 with @xmath160 ( figure [ light_curve_v598_pup_v1500_cyg ] ) and @xmath161 ( table [ light_curves_of_novae_ne ] ) .
the difference between co and neon nova models is only 0.2 mag .
therefore , this method for obtaining the distance modulus is rather robust against the ambiguity of chemical composition of a nova envelope .
the distance to v1500 cyg was estimated to be @xmath162 kpc from the nebular ( ejecta ) expansion parallax ( e.g. , * ? ? ?
the absorption was obtained to be @xmath163 by @xcite , so that the distance modulus to v1500 cyg is @xmath164 , being consistent with the value of 12.5 ( `` co nova 2 '' ) and 12.7 ( `` ne nova 2 '' ) estimated above based on our new method .
thus , the absolute magnitude estimates of v1500 cyg based on our free - free emission model light curves are consistent with the distance derived from the nebular expansion parallax method @xcite within an accuracy of 0.2 mag in the distance modulus of @xmath143 .
this in turn supports our assumptions and simplifications on free - free emission of novae .
the second example is gk per ( nova persei 1901 ) .
we assume that , when two novae have a similar decline rate ( or a similar timescale ) of free - free light curve , their brightnesses are the same during the period in which the two free - free light curves are overlapped .
figure [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ] shows such an example , a comparison between the light curves of gk per and v1500 cyg .
gk per has a decline timescale very similar to that of v1500 cyg . even in a typical transition phase ( from day @xmath165 until day @xmath166 )
, we are able to nicely fit our universal decline law with the observation , especially the bottom line of each oscillation until day @xmath152 as shown in the figure .
the top line connecting maxima of each oscillation deviates largely from our free - free model light curve .
this deviation is caused mainly by the increase in continuum flux itself .
the effect of strong emission lines starts from day @xmath152 , where the bottom line of each oscillation starts to deviate from our model light curve .
our best - reproducing model is a @xmath133 wd among other wd masses in table [ light_curves_of_novae_ne ] .
then we obtain the distance modulus of @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w}= 10.2 - 0.9 = 9.3\ ] ] from @xmath168 in figure [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ] and @xmath161 in table [ light_curves_of_novae_ne ] , where ff means the method of our free - free ( ff ) light curve fitting .
direct comparison of the brightness between gk per and v1500 cyg gives the difference of their distance moduli . from figure [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ]
, we have @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm gk~per } = + 3.4 , \label{distance_modulus_v1500cyg_gkper}\ ] ] where @xmath170 is the distance in units of pc and @xmath171 is the absorption in the @xmath71 band . with @xmath172 kpc and @xmath173 for v1500 cyg and @xmath174 kpc @xcite and @xmath175 @xcite for gk per
, we obtain @xmath176_{\rm v1500~cyg}= 12.5 , \label{distance_modulus_v1500_cyg}\ ] ] and @xmath97_{\rm gk~per } = 9.2 .
\label{distance_modulus_gk_per}\ ] ] therefore , the difference of @xmath177 is consistent with the difference in equation ( [ distance_modulus_v1500cyg_gkper ] ) within an accuracy of 0.1 mag and the distance modulus of equation ( [ distance_modulus_gk_per ] ) is also consistent with the result of the above ff method ( 9.3 mag ) within 0.1 mag .
thus , we confirm that the brightnesses of the two novae are the same within an uncertainty of @xmath178 mag .
this good agreement supports our assumption that two novae have the same brightness during their overlapping period if the two novae have a similar timescale as shown in figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] .
the third example is v1493 aql ( nova aquilae 1999 no.1 ) .
this nova shows a peculiar secondary peak ( figure [ all_mass_v1493_aql_v1500_cyg_x55z02o10ne03 ] ) , which is seen in all the bands , i.e. , @xmath71 , @xmath179 , and @xmath180 bands .
@xcite explained this kind of secondary peak by an additional energy source from strong magnetic field ( and rotation ) .
therefore , we regard that the nova light curves come back to a free - free emission light curve after the additional energy input stopped .
we see that , just after the secondary peak , the light curve sharply drops to merge into the light curve of v1500 cyg .
so we expect that v1493 aql has a similar decline rate to v1500 cyg , i.e. , our free - free model light curve of @xmath153 with `` ne nova 2 '' agrees well with the observation .
the effect of strong emission lines starts only from day @xmath152 , where the @xmath71 and visual magnitudes begin to depart from the @xmath180 magnitude . using a @xmath133 wd model in table [ light_curves_of_novae_ne ]
, we obtain the distance modulus of @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w}= 19.3 - 0.9 = 18.4\ ] ] from @xmath181 in figure [ all_mass_v1493_aql_v1500_cyg_x55z02o10ne03 ] and @xmath161 in table [ light_curves_of_novae_ne ] .
direct comparison of the brightness between v1493 aql and v1500 cyg gives the difference of their distance moduli as shown in figure [ all_mass_v1493_aql_v1500_cyg_x55z02o10ne03 ] , i.e. , @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v1493~aql } = -5.7.\ ] ] with @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , we obtain the distance modulus of @xmath167_{\rm lc } = \left [ 5 \log(d/10 ) + a_v \right]_{\rm v1493 aql } = 18.2 , \label{distance_absorption_v1493aql_v}\ ] ] where lc means the method of our direct light curve ( lc ) fitting .
the distance modulus of this nova was estimated to be @xmath183 @xcite , @xmath184 @xcite , @xmath185 @xcite , and @xmath186 @xcite , based on the mmrd relations .
the simple average values of these four observational estimates becomes @xmath187 , being consistent with our estimate of the ff and lc methods .
if we adopt @xmath188 @xcite , the distance is @xmath189 and @xmath190 kpc for ff and lc method , respectively .
these values are consistent with @xmath191 kpc obtained by @xcite based on the mmrd relations . in the present paper
, we proposed two new methods for obtaining distance modulus of a nova : one is fitting our free - free ( ff ) emission model light curve with the observation ( we call this ff method ) . the other is fitting the light curve ( lc ) of a distance - unknown nova with the other distance - well - known nova ( we call this lc method ) . from the above confirmations
, we may conclude that our new methods give a reasonable distance .
it should be addressed that our new methods for obtaining distance moduli of novae are applicable even when we missed the optical maxima of nova light curves . in section [ light_curve_of_ten_novae ]
, we apply these two methods to ten novae .
there are several timescales that characterize nova light curves ; @xmath16 is the time in units of day at optical maximum ( usually @xmath71 magnitude ) ; @xmath192 and @xmath3 are the times in units of day , in which @xmath71 magnitude decays by 2 and 3 mag , respectively , from its maximum ( from @xmath16 ) ; @xmath193 is the epoch when our free - free model light curves ( or optical @xmath64 , near infrared @xmath180 , @xmath194 , @xmath150 , @xmath151 light curves ) bend sharply as shown in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] , [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] , and [ light_curve_v598_pup_v1500_cyg ] ; @xmath22 is the epoch when optically thick winds stop ; and @xmath195 is the epoch when hydrogen shell - burning ends . @xcite and
@xcite calculated free - free model light curves for various wd masses with many sets of chemical composition .
they fitted their model light curves with observation and derived @xmath193 , @xmath22 , and @xmath195 for many novae .
table [ light_curve_parameters_novae ] summarizes such values of @xmath193 , @xmath22 , and @xmath195 for 15 classical novae , taken from our published papers as well as @xmath192 and @xmath3 for the same 15 novae , which are taken from literature .
figures [ t2t3_t_break][t2t3_t_h_burning ] show these five timescales for the 15 novae .
when the optically thick wind stops , the photospheric radius shrinks drastically and the photospheric temperature increases up to @xmath196 ev ( e.g. , * ? ? ?
* ) so that the supersoft x - ray flux emerges ( see appendix b for the photospheric radius which is shrinking with the temperature of the wd envelope during a supersoft x - ray phase ) .
we regard that @xmath22 is almost equal to the emerging time of supersoft x - ray , @xmath197 . after the hydrogen shell - burning stops ,
the wd envelope quickly cools down and the supersoft x - ray phase ends .
so we regard that @xmath195 is almost equal to the decay time of supersoft x - ray , @xmath198 . among these five timescales of @xmath192 , @xmath3 , @xmath193 , @xmath22 , and @xmath195 ,
the first three timescales , @xmath192 , @xmath3 , and @xmath193 , are much shorter than the other two , @xmath22 and @xmath195 .
llrrrrrr gk per 1901 & 7 & 13 & 1 & 73 & 182 & 382 & 5 + v1500 cyg 1975 & 2.9 & 3.6 & 2 & 70 & 180 & 380 & 8 + v1668 cyg 1978 & 12.2 & 24.3 & 3 & 110 & 280 & 720 & 8 + v1974 cyg 1992 & 16 & 42 & 1 & 96 & 250 & 600 & 8 + v382 vel 1999 & 6 & 12 & 4 & 49 & 120 & 220 & 9 + v2361 cyg 2005 & 6 & 8 & 5 & 69 & 169 & 340 & 5 + v382 nor 2005 & 12 & 18 & 5 & 73 & 182 & 382 & 5 + v5115 sgr 2005 & 7 & 14 & 5 & 60 & 145 & 280 & 5 + v378
ser 2005 & 44 & 90 & 5 & 257 & 858 & 2560 & 5 + v5116 sgr 2005#2 & 20 & 33 & 5 & 114 & 319 & 757 & 5 + v1188 sco 2005 & 7 & 21 & 5 & 44 & 110 & 190 & 5 + v1047 cen 2005 & 6 & 26 & 5 & 257 & 858 & 2560 & 5 + v476 sct 2005 & 15 & 28 & 6 & 100 & 260 & 590 & 5 + v1663 aql 2005 & 13 & 26 & 5 & 100 & 260 & 590 & 5 + v477 sct 2005#2 & 3 & 6 & 7 & 32 & 80 & 121 & 5 in what follows we introduce several relations between these five timescales , @xmath192 , @xmath3 , @xmath193 , @xmath22 , and @xmath195 . in figure
[ t2t3_t_break ] , we plot @xmath195 , @xmath22 , and @xmath193 against @xmath193 , which are theoretical ones calculated from our free - free model light curves @xcite .
these values depend on @xmath32 and the chemical composition , but are actually independent of @xmath199 .
this is because the difference in @xmath199 makes a large difference in the optical maximum but affects only the very early phase of nova outbursts as shown in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] and the difference in the epoch of optical maximum is so short ( a few to several days ) that it can be neglected compared with the timescales of @xmath195 , @xmath22 , and @xmath193 ( several tens of days or a few hundred days ) .
so we can derive relations among @xmath195 , @xmath22 , and @xmath193 from our models ( independently of observations ) as shown in figures [ t2t3_t_break][t2t3_t_h_burning ] .
once we determine @xmath193 of individual novae from observations , we can theoretically predict the duration of a luminous supersoft x - ray phase from figure [ t2t3_t_break ] . from figure [ t2t3_t_break ]
, we obtain approximate relations between @xmath193 , turn - on ( @xmath200 ) , and turnoff ( @xmath201 ) times as @xmath202 @xmath203 and @xmath204 for @xmath205 days .
it should be noted that the proportionality of @xmath22 to @xmath193 , i.e. , @xmath206 , is a direct result of our scaling law ( the universal decline law ) , because both @xmath193 and @xmath22 are during the optically thick wind phase which is the subject of universal decline law . then the @xmath22 is directly proportional to @xmath193 . on the other hand , the turnoff time of hydrogen shell burning , @xmath195 , is not proportional to @xmath193 or @xmath22 , because the duration of static hydrogen shell burning phase depends on hydrogen content @xmath33 ( amount of fuel ) of the envelope and wd mass ( temperature ) in a different way from the optically thick wind phase , in which the decay rate is determined mainly by wind mass - loss not by hydrogen burning rate ( see figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] ) .
therefore , equations ( [ t_break - t_h_burning ] ) and ( [ t_wind - t_h_burning ] ) are empirical laws obtained numerically , the power of which is not 1.0 but approximately 1.5 .
it is , however , not easy to obtain @xmath193 because nova light curves such as @xmath71 magnitudes are often heavily contaminated by strong emission lines in the later phase and this contamination easily clouds the break point of light curves as can be seen in figure [ light_curve_v598_pup_v1500_cyg ] . in the next subsection
, we derive relations between @xmath3 time and two of @xmath22 and @xmath195 times , because @xmath3 time is determined relatively easily from early optical observation .
llllllll v598 pup 2007#2 & ... & & & & & & + v382 vel 1999 & ... & 7.5 & 15 & @xmath207yes@xmath208 & & no & @xcite + v382 vel 1999 & ... & 4.5 & 9 & @xmath207yes@xmath208 & & @xmath207yes@xmath208 & @xcite + v382 vel 1999 & ... & 6 & 12 & yes & & yes & @xcite + v382 vel 1999 & ... & & 12.5 & & & yes & @xcite + v382 vel 1999 & ... & 6 & 10 & yes & & yes & @xcite + v4743 sgr 2002#3 & ... & 12 & 22 & yes &
& @xmath207yes@xmath208 & @xcite + v4743 sgr 2002#3 & ... & & 15 & & & @xmath207yes@xmath208 & @xcite + v4743 sgr 2002#3 & ... & 9 & 16 & yes & & yes & @xcite + v1281 sco 2007#2 & ... & & & & & & + v597 pup 2007#1 & ... & 2.5 & & & & & @xcite + v1494 aql 1999#2 & ... & 8 & 16 & @xmath207yes@xmath208 & no & no & @xcite + v1494 aql 1999#2 & ... & 7 & 23 & no & yes & @xmath207yes@xmath208 & @xcite + v1494 aql 1999#2 & ... & 6.1 & 15.8 & no & no & no & @xcite + v1494 aql 1999#2 & ... & 6 & 15 & no & no & no & @xcite + v1494 aql 1999#2 & ... & 6.6 & 16 & no & no & no & @xcite + v2467 cyg 2007 & ... & 7.3 & 15.1 & @xmath207yes@xmath208 & & no & @xcite + v2467 cyg 2007 & ... & 7.6 & 14.6 & yes & & no & @xcite + v2467 cyg 2007 & ... & 8 & & & & & @xcite + v5116 sgr 2005#2 & ... & 6.5 & 20.2 & no & & no & @xcite + v5116 sgr 2005#2 & ... & 18 & 45 & no & & no & @xcite + v5116 sgr 2005#2 & ... & 20 & 33 & yes & & yes & @xcite + v574 pup 2004 & ... & 37 & 85 & no & no & no & @xcite + v574 pup 2004 & ... & 13 & 58 & no & no & no & @xcite + v458 vul 2007 & ... & 8 & 31 & no & no & & @xcite + v458 vul 2007 & ... & 7 & 15 & @xmath207yes@xmath208 & no & & @xcite + v458 vul 2007 & ... & 7 & 18 & no & no & & @xcite + + v1974 cyg 1992 & ... & 22 & 48 & no & no & no & @xcite + v1974 cyg 1992 & ... & 16 & 42 & no & no & no & @xcite + v1668 cyg 1978 & ... & 12.2 & 24.3 & yes & &
& @xcite + v1500 cyg 1975 & ... & 2.9 & 3.6 & & & & @xcite it is very useful if we can predict @xmath197 ( @xmath209 ) and @xmath198 ( @xmath210 ) from @xmath3 time , because @xmath3 is popular and widely used to define the nova speed class . in this subsection , we derive relations between @xmath3 and @xmath197 , and between @xmath3 and @xmath198 . as already explained in the previous subsection , @xmath193 , @xmath22 , and @xmath195
are practically independent of the initial envelope mass , @xmath199 , but @xmath3 time depends on the peak brightness of an individual nova . as shown in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] , the initial envelope mass @xmath199 is larger , the optical maximum is brighter .
therefore , @xmath192 and @xmath3 depend on @xmath199 .
this is the reason why we can not theoretically derive @xmath192 and @xmath3 .
we need information of maximum brightness to specify @xmath3 time for individual novae .
now we go back to the well - observed v1668 cyg which we regard as a typical classical nova .
thus , we adopt point b in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] as a typical ( an average ) case of maximum brightness . in this case ,
@xmath211 and @xmath212 days are derived from figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] . then , we have @xmath213 ( see table [ fitting_t2_t3_tb_txonoff ] for v1668 cyg timescales ) .
using the universal decline law , we obtain a relation between @xmath3 and @xmath193 as @xmath214 for @xmath2 days , where the upper and lower bounds are corresponding to @xmath215 mag brighter / darker ( more massive / less massive initial envelope mass ) points than that of point b. this relation is indicated by a red thick solid line in figure [ t2t3_t_break ] together with the upper and lower red dashed lines that denote upper / lower values ( @xmath216 ) of equation ( [ t_3-t_break ] ) .
there is a relation between @xmath192 and @xmath3 derived from the scaling law of our universal decline law , which is also well used in literature . @xcite
have already derived @xmath217 for the optical light curves that follow the free - free model light curve with @xmath2 days .
this relation has been empirically known @xcite and theoretically confirmed by @xcite based on the universal decline law with a slope of @xmath218 as shown in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] and [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] . inserting the central value of equation ( [ t_2-t_3 ] ) into equation ( [ t_3-t_break ] )
, we obtain relations between @xmath193 and @xmath192 relations by blue solid and dashed lines as shown in figure [ t2t3_t_break ] .
figure [ t2t3_t_break ] also shows @xmath192 and @xmath3 times for individual novae in table [ light_curve_parameters_novae ] .
four novae indicated by colors are far from the relation of equation ( [ t_3-t_break ] ) , but there is a special reason in each nova as described in the figure caption .
the other novae reasonably follow the above relation with the reasonable scatter corresponding to @xmath215 mag brighter / darker of the peak brightness .
this supports that v1668 cyg has an average @xmath199 . in the same way
, we derive a relation between @xmath3 and @xmath22 from equations ( [ t_break - t_wind ] ) and ( [ t_3-t_break ] ) , i.e. , @xmath219 for @xmath2 days . figure [ t2t3_t_wind ] shows the five timescales against @xmath22 and we also see a reasonable fit of our relation ( equation ( [ t_3-t_wind ] ) ) with the observation . from equations ( [ t_break - t_h_burning ] ) and ( [ t_3-t_break ] )
, we derive a relation between @xmath3 and @xmath195 , i.e. , @xmath220 for @xmath2 days .
figure [ t2t3_t_h_burning ] shows the five timescales against @xmath195 and we can also see a reasonable fit of equation ( [ t_3-t_h_burning ] ) with the observation . however , it is not always easy to determine @xmath3 ( or @xmath192 ) time from observation .
in fact , table [ t2_t3_txonoff ] shows nova timescales @xmath192 and @xmath3 of 11 novae taken from literature , which are sometimes very different from author to author , and sometimes very different from our estimates in table [ fitting_t2_t3_tb_txonoff ] .
therefore , we propose a more robust way . instead of observational face values of @xmath192 and @xmath3 , we will use `` intrinsic '' @xmath192 and @xmath3 , which are obtained from our model light curves fitted with observation .
for example , we use @xmath211 and @xmath212 days for v1668 cyg obtained from figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] instead of observational face values .
this `` intrinsic '' @xmath192 and @xmath3 times are especially useful when the observational data deviate from our model light curve near the peak . in what follows
, we will see how to determine `` intrinsic '' @xmath3 .
figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] explained how to select our best model light curve from early phase observations .
it also shows that three different wd mass ( 0.85 , 0.95 , and @xmath119 ) models can reproduce the early phase light curve and we can hardly select one only from the early phase optical and ir light curves .
however , this uncertainty of mass determination is not a problem in the estimation of `` intrinsic '' @xmath3 time , because these three model light curves , of course , give similar values of @xmath3 time ( @xmath221 days ) along our model light curves . using relations of equations ( [ t_break - t_wind])([t_3-t_h_burning ] ) , we can predict the period of a supersoft x - ray phase of individual novae from @xmath3 ( or @xmath192 , @xmath193 ) time . in section [ light_curve_of_ten_novae ] , we will show that our formulae are much better fitted with observation if we use intrinsic @xmath3 time than if we use face values of @xmath3 .
therefore , in what follows in this paper , we use intrinsic @xmath3 time in order to predict @xmath200 and @xmath201 from equations ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) .
we assumed here that a nova has a typical @xmath199 , i.e. , has a peak brightness similar to that of v1668 cyg ( i.e. , point b in the rescaled @xmath222@xmath79 plane , figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) .
if the rescaled peak magnitude is much brighter / darker than that of v1668 cyg ( for example , points a / c in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) , x - ray on and off times , @xmath200 and @xmath201 , can deviate from the prediction of equations ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) . in this sense ,
our prediction formulae ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) are statistical ones .
the maximum magnitude rate of decline ( mmrd ) relation is often used to estimate the distance to novae ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this section ,
we derive theoretical mmrd relations based on our universal decline law .
figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] and [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] clearly shows a trend that a more massive wd has a brighter maximum magnitude ( smaller @xmath223 ) and a faster decline rate ( smaller @xmath3 time ) .
the relation between @xmath3 and @xmath223 for novae is called `` the maximum magnitude rate of decline '' ( mmrd ) relation . in order to derive theoretical mmrd relations , we go back to figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] and utilize the universal decline law starting from point b ( a typical case of nova envelope mass ) . we have already shown that all the free - free emission model light curves for different wd masses overlap with each other ( they converge to @xmath61 wd in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] against v1668 cyg data ) when they are squeezed in the direction of time by a factor of @xmath66 , data of which are tabulated in table [ light_curves_of_novae_co ] .
now we assume that all the free - free emission model light curves reaches the same maximum point , like point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
in such a case , the apparent maximum brightness @xmath224 of each light curve with different wd masses is expressed as @xmath225 .
the @xmath3 time of each model light curve with different wd masses is squeezed to be @xmath226 .
eliminating @xmath66 from these two relations , we have @xmath227 since point b corresponds to the optical peak of v1668 cyg ( @xmath102 in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) , we have already known @xmath228 days and @xmath229 . substituting these values of @xmath230 into equation ( [ theoretical_apparent_mmrd_relation ] ) , we have @xmath231 this is our theoretical apparent mmrd ration for a starting point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
next , using the distance modulus of @xmath143 , we obtain our theoretical mmrd relation as @xmath232 substituting @xmath233 , the distance modulus of v1668 cyg , into equation ( [ theoretical_mmrd_relation ] ) , we have @xmath234 figure [ max_t3_scale ] shows this theoretical mmrd relation for novae that has a typical ( average ) @xmath199 like v1668 cyg . in the right axis
we convert the apparent magnitude to the absolute magnitude by using the distance modulus of v1668 cyg .
point b in figure [ max_t3_scale ] corresponds to v1668 cyg ( @xmath102 : @xmath61 , in which nova light curves start from point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) .
the other wd mass cases starting from the same point b in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] have different @xmath235 .
changing @xmath66 ( or @xmath32 ) , we have a line passing through point b in figure [ max_t3_scale ] .
this is our theoretical mmrd relation , the parameter of which is @xmath32 ( or @xmath66 ) .
we also show the equivalent parameter , stretching factor @xmath66 ( against point b ) , in the upper axis of the same figure .
the black thick solid line passing through point b shows quite a good agreement with the central blue and magenta lines , which are empirical mmrd relations obtained from observation as will be explained bellow .
if the initial envelope mass @xmath199 is larger or smaller than that of point b ( which is an average value , see figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) , we obtain different mmrd relations . in figure
[ max_t3_scale ] , we show five points of green filled squares corresponding to different initial envelope mass of @xmath38 . point a @xmath236 days , @xmath237 corresponds to a larger envelope mass at ignition ( point a in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) . the black solid line passing through point a is other @xmath238@xmath239 relation for a much larger initial envelope mass of @xmath199 ( at point a in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) , i.e. , @xmath240 on the other hand , the black solid line passing through point c @xmath241 days , @xmath242 is also different @xmath238@xmath239 relation for a much smaller initial envelope mass ( at point c in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) , i.e. , @xmath243 for a given wd mass , @xmath199 is larger ( smaller ) for a smaller ( larger ) accretion rate .
points a , b , and c correspond to different ignition masses for a given wd mass .
when the accretion rate is much smaller than that for point b , the brightness reaches , for example , point a at maximum and then we have a shorter @xmath3 time ( see figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] ) . in the case of a larger accretion rate than that for point b
, the brightness goes up only to point c at maximum and we have a longer @xmath3 time .
so , even for a given @xmath32 and chemical composition of the envelope , we have different @xmath3 time , depending on @xmath199 .
in other words , a variation of mass accretion rates from an average value makes a divergence ( or width ) in the mmrd relation for individual novae .
the five points including points a , b , and c are calculated for v1668 cyg model light curves , all of which correspond to the 0.95 @xmath80 wd of `` co nova 2 '' ( @xmath244 ) .
the green solid line connecting these five points represents the @xmath245@xmath239 relation for 0.95 @xmath80 wd of `` co nova 2 . ''
we then obtain other @xmath245@xmath239 lines for other wd masses as shown by green lines ( attached numbers represent wd masses ) in figure [ max_t3_scale ] . in the case of neon novae
, we also obtain similar lines , which are also shown in figure [ max_t3_scale ] .
these lines are almost overlapped to those of `` co nova 2 '' but different wd masses ( blue attached numbers for wd masses ) .
our mmrd ( @xmath245@xmath239 ) relation passing through point b is in a good agreement with two well - known mmrd relations empirically obtained : one is kaler - schmidt s law @xcite , i.e. , @xmath246 and the other is della valle & livio s ( 1995 ) law , i.e. , @xmath247 where @xmath223 is the absolute @xmath71 magnitude at maximum and we use equation ( [ t_2-t_3 ] ) to calculate @xmath192 from @xmath3 in equation ( [ della - valle - livio - law ] ) in figure [ max_t3_scale ] .
kaler - schmidt s law is denoted by a blue solid line with two attendant blue solid lines corresponding to @xmath215 mag brighter / darker cases .
della valle & livio s law is indicated by magenta solid lines .
these two well - known empirical mmrd relations are very close to our theoretical mmrd relation at average .
we also add observational mmrd points ( @xmath245 and @xmath3 ) for individual novae , data of which are taken from table 5 of @xcite , although their @xmath3 are not intrinsic but observational face values .
it is clearly shown that a large scatter of individual points ( red filled circles ) from the two empirical mmrd relations falls into between upper / lower cases of our mmrd relations for the largest / smallest initial envelope mass of @xmath199 .
this simply means that there is a second parameter to specify the mmrd relation for individual novae .
the main parameter is the wd mass represented by stretching factor @xmath66 .
the second parameter is the initial envelope mass ( or the mass accretion rate to the wd ) .
this second parameter can reasonably explain the scatter of individual novae from the empirical mmrd relations ever proposed .
it is interesting to place the position of the recurrent nova rs oph in this diagram .
rs oph locates in the smallest corner of @xmath3 time and a relatively darker level of @xmath223 , near the @xmath223@xmath239 relation passing through point c. this indicates a much smaller envelope mass at the optical maximum , suggesting that the mass accretion rate to the wd is very high .
this situation is very consistent with the total picture of recurrent novae ; a very massive wd close to the chandrasekhar mass and a high mass accretion rate to the wd ( e.g. , * ? ? ?
* ) . in this figure
, we adopt @xmath248 and @xmath249 days with the distance of @xmath250 kpc @xcite , absorption of @xmath251 @xcite , @xmath252 @xcite , and @xmath249 days from optical light curve fitting with our free - free model light curves @xcite .
in this section , we examine ten classical novae which we have not yet analyzed in our series of papers , i.e. , v598 pup , v382 vel , v4743 sgr , v1281 sco , v597 pup , v1494 aql , v2467 cyg , v5116 sgr , v574 pup , and v458 vul , in the order of increasing timescale .
all are supersoft x - ray detected novae , so we can check whether or not our prediction formulae of a supersoft x - ray phase work on these novae .
llllllllll v598 pup 2007#2 & ... & & & 40 & 90 & 150 & & 1.28 & ne nova 2 + v382 vel 1999 & ... & 6.6 & 12.4 & 52 & 120 & 220 & 32 & 1.23 & ne nova 2 + v4743 sgr 2002#3 & ... & 9.4 & 17 & 70 & 180 & 380 & 36 & 1.15 & ne nova 2 + v1281 sco 2007#2 & ... & 9.5 & 16.7 & 78 & 190 & 425 & & 1.13 & ne nova 2 + v597 pup 2007#1 & ... & 8.7 & 16.5 & 74 & 200 & 430 & 52 & 1.2 & ne nova 3 + v1494 aql 1999#2 & ... & 11.1 & 21.2 & 96 & 260 & 670 & 66 & 1.13 & ne nova 3 + v1494 aql 1999#2 & ... & 11.7 & 22.1 & 102 & 260 & 670 & 72 & 1.06 & ne nova
2 + v1494 aql 1999#2 & ... & 14.7 & 26.6 & 120 & 285 & 675 & 95 & 0.92 & co nova 2 + v2467 cyg 2007 & ... & 14.3 & 25.3 & 110 & 290 & 765 & 68 & 1.11 & ne nova 3 + v2467 cyg 2007 & ... & 14.8 & 27.9 & 113 & 290 & 765 & 75 & 1.04 & ne nova 2 + v2467 cyg 2007 & ... & 17.6 & 31.7 & 135 & 310 & 750 & 96 & 0.90 & co nova 2 + v5116 sgr 2005#2 & ... & 16.7 & 30.4 & 127 & 355 & 1000 & 91 & 1.07 & ne nova 3 + v5116 sgr 2005#2 & ... & 16.2 & 30.5 & 134 & 350 & 995 & 100 & 1.0 & ne nova 2 + v5116 sgr 2005#2 & ... & 20.0 & 35.8 & 169 & 390 & 1010 & 130 & 0.85 & co nova 2 + v574 pup 2004 & ... & 18.6 & 35.0 & 144 & 390 & 1140 & & 1.05 & ne nova 3 + v458 vul 2007 & ... & 20.4 & 37.2 & 145 & 415 & 1250 & 62 & 0.93 & co nova 4 + + v1974 cyg 1992 & ... & 16 & 26 & 108 & 250 & 600 & 90 & 1.05 & ne nova 2 + v1668 cyg 1978 & ... & 14.4 & 26 & 100 & 280 & 720 & 87 & 0.95 & co nova 3 + v1500 cyg 1975 & ... & 7.2 & 13 & 70 & 180 & 380 & 50 & 1.15 & ne nova 2 llllllllll v598 pup 2007#2 & ... & 1.28 & & & & & 0.31 & 0.34 & ne nova 2 + v382 vel 1999 & ... & 1.23 & 0.31 & 0.1461 & 1.35 & 0.68 & 0.43 & 0.48 & ne nova 2 + v4743 sgr 2002#3 & ... & 1.15 & 0.70 & 0.2799 & 2.2 & 0.93 & 0.69 & 0.78 & ne nova 2 + v1281 sco 2007#2 & ... & 1.13 & & & & & 0.92 & 1.02 & ne nova 2 + v597 pup 2007#1 & ... & 1.2 & 0.22 & 0.1112 & 1.1 & 0.58 & 1.06 & 1.14 & ne nova 3 + v1494 aql 1999#2 & ... & 1.13 & 0.28 & 0.1346 & 1.24 & 0.62 & 1.04 & 1.12 & ne nova 3 + v1494 aql 1999#2 & ... & 1.06 & 0.28 & 0.1346 & 1.22 & 0.60 & 1.11 & 1.26 & ne nova 2 + v1494 aql 1999#2 & ... & 0.92 & 0.28 & 0.1346 & 1.17 & 0.57 & 1.75 & 2.00 & co nova 2 + v2467 cyg 2007 & ... & 1.11 & 0.35 & 0.1596 & 1.40 & 0.67 & 1.08 & 1.23 & ne nova 3 + v2467 cyg 2007 & ... & 1.04 & 0.35 & 0.1596 & 1.38 & 0.65 & 1.16 & 1.34 & ne nova 2 + v2467 cyg 2007 & ... & 0.90 & 0.35 & 0.1596 & 1.33 & 0.62 & 1.67 & 1.90 & co nova 2 + v5116 sgr 2005#2 & ... & 1.07 & 0.25 & 0.1238 & 1.15 & 0.58 & 1.25 & 1.43 & ne nova 3 + v5116 sgr 2005#2 & ... & 1.00 & 0.25 & 0.1238 & 1.13 & 0.56 & 1.30 & 1.53 & ne
nova 2 + v5116 sgr 2005#2 & ... & 0.85 & 0.25 & 0.1238 & 1.08 & 0.53 & 2.17 & 2.52 & co nova 2 + v574 pup 2004 & ... & 1.05 & & & & & 1.25 & 1.46 & ne nova 3 + v458 vul 2007 & ... & 0.93 & 0.74 & 0.5895 & 3.51 & 1.40 & 1.26 & 1.53 & co nova 4 + + v1974 cyg 1992 & ... & 1.05 & 0.15 & 0.0813 & 0.85 & 0.44 & 1.0 & 1.17 & ne nova 2 + v1668 cyg 1978 & ... & 0.95 & 0.29 & 0.1384 & 1.21 & 0.58 & 1.1 & 1.3 & co nova 3 + v1500 cyg 1975 & ... & 1.15 & 0.29 & 0.1396 & 1.28 & 0.64 & 0.71 & 0.8 & ne nova 2 v598 puppis ( nova puppis 2007 no.2 ) was serendipitously discovered by the _ xmm - newton _ slew survey @xcite .
@xcite subsequently found an optically bright counter part of this x - ray object .
the spectra obtained on november 16.34 ut ( day @xmath253 ) reveal numerous features suggestive of a nova that already entered the auroral phase ( see also * ? ? ? * ) .
analysis of all sky automated survey ( asas ) archive showed that the @xmath71 magnitude rose to @xmath254 from @xmath255 between june 1.966 ut and june 5.968 ut @xcite .
since we can not specify the exact date of the outburst , we assume here that the outburst day is @xmath256 jd 2454253.0 ( 2007 june 1.5 ut ) .
the x - ray satellite _ swift _ detected x - ray flux from v598 pup , which declined rapidly from @xmath257 ( day @xmath258 ) to @xmath259 ( day @xmath166 ) and then to @xmath260 erg s@xmath261 @xmath262 ( at day @xmath253 after the outburst ) @xcite as shown in figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] .
this quick drop in the x - ray flux clearly showed that hydrogen burning on the wd ended around 150 days after the outburst .
figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] shows observed optical light curves of v598 pup as well as the x - ray data .
unfortunately there are only two observational points available before 77 days after the outburst .
to understand the characteristic properties of v598 pup light curves , we add observational points of v1500 cyg that show a very similar light curve in the later phase .
we can see that two points of @xmath71 mag in the early phase are just on our free - free model light curve but the @xmath71 data after day @xmath154 are those in the nebular phase and already deviated from the expected free - free light curve .
this deviation often occurs in novae when @xmath71 magnitude is contaminated by strong emission lines such as [ ] in the nebular phase ( see * ? ? ? * for emission lines of v598 pup ) , as already discussed in @xcite and @xcite .
thus we regard that the light curve of v598 pup decays very similarly to that of v1500 cyg , which follows our universal decline law from day @xmath126 to day @xmath154 .
no abundance determinations are available for v598 pup so far .
we here assume two typical cases , one is for oxygen - neon novae of `` ne nova 2''and the other is for carbon - oxygen novae of `` co nova 2 '' as listed in table [ chemical_composition ] .
the emergence / decay times of supersoft x - ray are a good indicator of the wd mass @xcite .
we have calculated several models for different wd masses with these two chemical compositions . among the models of @xmath263 , @xmath264 , and @xmath265 ,
we have selected @xmath266 as a best - reproducing model for `` ne nova 2 , '' but , @xmath267 for `` co nova 2 '' as a best - reproducing one among @xmath268 , @xmath269 , and @xmath133 , which are not shown in the figure .
the best models are selected by careful eye fitting with the supersoft x - ray data because there are only three points .
the optical fitting can not be used for the selection because all of three model light curves go through the two observational @xmath71 points as seen in figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] .
we define the `` error '' in the wd mass determination ( @xmath270 ) from larger and smaller wd masses , @xmath263 and @xmath265 , safely cover the decay phase of supersoft x - ray flux for the fixed chemical composition .
the estimated wd mass of @xmath110 for `` co nova 2 '' exceeds the upper limit mass for co wds born in a binary , i.e. , @xmath271 @xcite , so we regard that the wd of v598 pup is very likely an onemg wd .
therefore , we adopt the @xmath272 wd of `` ne nova 2 '' as a best model , which has @xmath273 , @xmath274 , and @xmath275 days after the outburst ( see figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] ) . we have also calculated the ejecta mass of @xmath276 , which is lost by the optically thick wind ( see , e.g. , * ? ? ?
* ; * ? ? ?
* for more detail of ejecta mass calculation ) .
however , it should be noted that @xmath277 could be the subject to significant uncertainty because we missed the real optical maximum .
various physical parameters of novae thus obtained are summarized in tables [ fitting_t2_t3_tb_txonoff ] and [ physical_properties_of_novae ] .
we are not able to determine @xmath192 and @xmath3 from the light curve because of lack of visual magnitudes around the optical peak .
the mass determination of a wd depends weakly on the chemical composition of the envelope , especially , on the hydrogen content @xmath33 @xcite .
we have estimated this dependency from two other chemical composition models , i.e. , `` ne nova 1 '' and `` ne nova 3 '' in table [ chemical_composition ] .
we similarly obtain a best - reproducing model of @xmath267 for @xmath278 ( `` ne nova 1 '' ) and @xmath279 for @xmath280 ( `` ne nova 3 '' ) .
the resultant dependency can be approximated as @xmath281 for v598 pup . here
, we have already obtained @xmath282 . therefore
, we may conclude that the wd mass is @xmath283 for a typical hydrogen content between @xmath284 for v598 pup .
the wd mass estimate depends also on the cno abundance , but the difference among various @xmath35 models in table [ chemical_composition ] is smaller than the difference among various hydrogen content ( @xmath33 ) models ( see * ? ? ?
* for more detail of @xmath33 dependency ) .
the distance to v598 pup can be estimated by two methods even if the optical maximum is not available .
the first one is to use the absolute magnitude of our free - free model light curves in table [ light_curves_of_novae_ne ] .
we use our best - reproducing model in figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] and measure the brightness at the end of the free - free ( ff ) light curve , @xmath285 , and obtain the distance modulus to v598 pup as @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 11.8 - 0.1 = 11.7.\ ] ] here we adopt @xmath286 from figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] and use the absolute magnitude of @xmath287 at the end point of @xmath272 wd model light curve , which is calculated by interpolation between the @xmath265 and @xmath131 wd models in table [ light_curves_of_novae_ne ] .
we call this `` ff method '' as introduced in section [ checking_absolute_magnitude ]
. it should be addressed that the ff method is rather robust and gives @xmath288 for @xmath131 wd or @xmath289 for @xmath265 wd even if we miss a best model .
we also emphasize that this ff method does not require the optical maximum like the mmrd relations , so it can be used even if the optical maximum is not available like v598 pup .
the next method is to use direct comparison with the other nova light curve the distance modulus of which is already known .
we have already done this method in section [ checking_absolute_magnitude ] ( see figures [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] , [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ] , and [ all_mass_v1493_aql_v1500_cyg_x55z02o10ne03 ] ) .
the direct comparison between v1500 cyg and v598 pup in figure [ all_mass_v598_pup_v1500_cyg_x55z02o10ne03 ] gives the difference of distance moduli of these two novae , i.e. , @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v598~pup } = 0.5.\ ] ] when we adopt @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , we obtain @xmath167_{\rm lc } = \left [ 5 \log(d/10 ) + a_v \right]_{\rm v598~pup } = 12.0 , \label{distance_absorption_v598pup_lc}\ ] ] where lc means the direct light curve ( lc ) fitting .
it should be again addressed that the lc method does not need the optical maximum like the mmrd relations , so it can be used even if we missed the optical maximum like v598 pup .
the difference of 0.3 mag in the distance modulus estimation can be understood as follows .
these two nova light curves look very similar in the later phase but the fitting mass is different between @xmath272 ( v598 pup ) and @xmath133 ( v1500 cyg ) .
the absolute magnitude at 20 days , for example , is about 0.4 mag brighter in the @xmath133 wd model than in the @xmath265 as shown in figure [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] .
we have experienced that the same thing happened when we use fitting only of optical light curves as shown in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] . in this figure
, we obtain @xmath142 mag larger @xmath143 when the fitting mass is missed by @xmath141 , respectively . in this sense
, the lc method may have an error of 0.4 mag or so even if two nova light curves look very similar in the later phase .
the above two estimates are listed in table [ distance_moduli_of_novae ] , being roughly consistent with each other , i.e. , @xmath290 ( averaged with equal weight ) .
these two distance moduli give the distance of @xmath291 kpc for @xmath292 @xcite .
this distance is consistent with various estimates discussed by @xcite .
v382 velorum ( nova velorum 1999 ) was discovered independently by williams and gilmore @xcite .
the nova reached @xmath293 at maximum on 1999 may 23 and then decayed to @xmath294 @xcite as shown in figure [ all_mass_v382_vel_v1500_cyg_x55z02o10ne03 ] .
v382 vel is a fast nova and identified as a neon nova @xcite .
we here adopt a set of chemical composition of a typical neon nova , `` ne nova 2 '' in table [ chemical_composition ] , based on the composition analyses by @xcite and by @xcite ( see table 1 of * ? ? ?
we also assume that the outburst day is @xmath256jd 2451319.0 ( 1999 may 20.5 ut ) from information of iau circulars 7176 and 7184 , i.e. , @xmath295 on may 20.57 ut , @xmath296 on may 21.62 ut , @xmath297 on may 21.725 ut , and @xmath298 on may 22.396 ut .
supersoft x - rays were detected by _
bepposax _ about six months after the outburst @xcite .
after that , the flux decayed rapidly as shown in figure [ all_mass_v382_vel_v1500_cyg_x55z02o10ne03 ] .
the _ chandra _
observations by @xcite and by @xcite suggested that the hydrogen burning ended around 220 days after the outburst .
the theoretical supersoft x - ray phase is very sensitive to the wd mass whereas it depends weakly on the chemical composition of the envelope .
our x - ray light curve fitting suggests that 1 ) the wd mass is as massive as @xmath299 and 2 ) the supersoft x - ray phase started from day @xmath300 .
we have also estimated the ejecta mass of @xmath301 ( table [ physical_properties_of_novae ] ) for the @xmath302 wd model .
@xcite derived the ejecta mass of @xmath303 from the flux of h@xmath304 emission , which is consistent with our estimate .
our model of @xmath302 wd shows @xmath305 , @xmath306 , and @xmath307 days after the outburst ( see figure [ all_mass_v382_vel_v1500_cyg_x55z02o10ne03 ] ) .
since the @xmath71 magnitude reached its maximum @xmath308 days after the outburst , we obtain `` intrinsic '' @xmath309 and @xmath310 days along our model light curve . for this nova ,
since the observed data decays almost along our model light curves , the `` face '' values of @xmath192 and @xmath3 are practically the same as those of our `` intrinsic '' @xmath192 and @xmath3 .
it should be noted here that the assumed epoch of @xmath256 jd 2451319.0 ( 1999 may 20.5 ut ) hardly affect our estimation of @xmath192 and @xmath3 times because we do not use the value at @xmath15 .
usually we move the model light curve `` back and forth '' and `` up and down '' against the observed value as shown in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] , and this process simply means that we fit our model light curve with the decay phase after maximum . in figure [ all_mass_v382_vel_v1500_cyg_x55z02o10ne03 ] , we add @xmath71 and @xmath64 light curves of v1500 cyg for comparison . from the difference between @xmath71 and @xmath64 bands
, we can roughly evaluate the influence of strong emission lines that contribute extra flux to @xmath71 magnitude above free - free continuum flux . comparing the light curves of v1500 cyg
, we regard that v382 vel light curve is contaminated by strong emission lines from @xmath140 days after the outburst ; all the @xmath71 , @xmath179 , and @xmath180 magnitudes begins to diverge from each other and start to deviate from our universal decline law ( thick solid line ) .
the wd mass estimate depends weakly on the chemical composition .
changing hydrogen content @xmath33 , we obtain a best - reproducing model of @xmath311 for @xmath278 ( `` ne nova 1 '' ) and @xmath312 for @xmath280 ( `` ne nova 3 '' ) .
thus , we obtain the dependency of the wd mass on @xmath33 , i.e. , @xmath313 for v382 vel . this relation on the wd mass is a bit steeper than that for v598 pup
. we may also conclude that the wd mass is @xmath314 for a typical hydrogen content between @xmath284 for v382 vel , which corresponds to the diversity in the composition determination between @xmath315 @xcite and @xmath316 @xcite .
the distance modulus to v382 vel can be calculated in the same two ways as those in the previous subsection ( v598 pup ) .
first , the ff method gives @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 11.9 - 0.5 = 11.4,\ ] ] for the @xmath302 wd model .
even if we use two other models of @xmath265 and @xmath110 wds in figure [ all_mass_v382_vel_v1500_cyg_x55z02o10ne03 ] , this method gives a similar result of @xmath317 for @xmath265 , or @xmath318 for @xmath110 wd .
second , the lc method also gives us the absolute brightness , i.e. , @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v382~vel } = 1.0 .
\label{distance_v1500cyg_v382vel}\ ] ] with @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , the above equation ( [ distance_v1500cyg_v382vel ] ) yields @xmath319_{\rm lc } = \left [ 5 \log(d/10 ) + a_v \right]_{\rm v382~vel } = 11.5 . \label{distance_absorption_v382vel_lc}\ ] ] kaler - schmidt s law with @xmath310 days and della valle & livio s law with @xmath309 days and @xmath320 @xcite give @xmath321_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 11.6 ~(11.4),\ ] ] where the value outside parenthesis is calculated from kaler - schmidt s law of equation ( [ kaler - schmidt - law ] ) and the value in parenthesis is calculated from della valle & livio s law of equation ( [ della - valle - livio - law ] ) .
these two values are consistent with the ff and lc results .
these four values give a distance modulus of @xmath322 for these four different estimations with equal weight and a distance of @xmath323 kpc for @xmath324 @xcite .
these values are consistent with the distance of @xmath325 kpc derived from the maximum magnitude vs. rate of decline ( mmrd ) relation @xcite . on the other hand
, @xcite obtained a bit larger distance of @xmath326 kpc , assuming that the uv flux of v382 vel is equal to that of v1974 cyg .
they took a distance of 3.1 kpc to v1974 cyg but this value is much larger than a reasonable one of 1.8 kpc @xcite . if we take the distance of 1.8 kpc instead of 3.1 kpc to v1974 cyg , shore et al.s method gives a much shorter distance of 1.5 kpc to v382 vel , which is consistent with our estimates .
finally , we introduce some characteristic features concerning binary nature , that is , the emergence of the companion from the wd photosphere .
we define this characteristic time as the emergence time of the companion , i.e. , @xmath327 .
the orbital period of @xmath328 days ( 3.51 hr ) was derived by @xcite from the orbital modulations with a full amplitude of @xmath329 mag while a bit longer orbital period of @xmath330 days was obtained by @xcite .
so , we estimate the epoch when the companion emerges from the nova envelope . if the mass of the donor star can be estimated from warner s ( 1995 ) empirical formula , i.e. , @xmath331 we have @xmath332 , which corresponds to the separation of @xmath333 for @xmath334 , and the effective roche lobe radius of the primary component ( wd ) of @xmath335 .
it is about @xmath336 days when the photospheric radius of the nova envelope shrinks to near the orbit ( when @xmath337 ) .
this becomes @xmath338 days if we take the orbit plus the companion s radius ( when @xmath339 ) .
hard x - ray ( @xmath340 kev ) flux of v382 vel was detected by _
asca _ and _ rxte _ @xcite about 20 days after the outburst ( day 18 with _
bepposax _ by * ? ? ?
* ) and reached maximum at day @xmath341 and stayed at the same level and then began to decrease at day @xmath342 .
@xcite interpreted the origin of hard x - rays by a model of internal shocks @xcite .
this shock may be formed by collision between two ejecta shells as explained by @xcite , but there is another possibility that the shock is formed by collision between nova winds ( optically thick winds ) and the companion star ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if it is the second case , the hard x - ray emergence should be coincident with the emergence of the companion from the wd photosphere because hard x - ray is probably absorbed deep inside the nova photosphere .
the hard x - ray emergence time of @xmath343 days is roughly consistent with @xmath344 days . on the other hand
, the hard x - ray flux should be declining as the wind mass - loss rate is decreasing .
this corresponds to the break point of free - free emission light curve , because the wind mass - loss rate is rapidly decreasing after the break point of @xmath305 days .
therefore , the decrease in the x - ray flux at day 50 is also consistent with our wind model .
after @xmath193 , the hard x - ray becomes weak but lasts about 120 days after the outburst , in other words , until the optically thick wind stopped at @xmath306 days .
v4743 sgr ( nova sagittarii 2002 no.3 ) was discovered by haseda on 2002 september 20.431 ut ( jd 2452537.931 ) at mag about 5.0 @xcite .
@xcite reported a prediscovery magnitude of mag 5.5 on september 18.465 ut .
this object was not detected by brown @xcite on september 9.6 ut ( limiting mag about 12.0 ) , so we here assume that @xmath256 jd 2452533.0 ( 2002
september 15.5 ut ) is the outburst day .
bright supersoft x - ray phases were reported at day @xmath345 ( 2003 march 19.396 ut ) by @xcite ( see also * ? ? ? * ) and at day @xmath346 ( 2003 april 4.921 ut ) by @xcite .
@xcite reported that the supersoft x - ray flux has already declined on 2004 september 30.783 ut ( day @xmath347 ) by a factor of 1000 .
figure [ all_mass_v4743_sgr_v1500_cyg_x55z02o10ne03 ] shows the optical and supersoft x - ray light curves .
the supersoft x - ray phase started at least @xmath345 days after the outburst and ended at day @xmath348 .
the model of @xmath133 wd is a best one among three 1.2 , 1.15 , and @xmath349 wds , as shown in figure [ all_mass_v4743_sgr_v1500_cyg_x55z02o10ne03 ] .
here we assume a chemical composition of typical neon novae , `` ne nova 2 . ''
the @xmath133 wd model has @xmath350 , @xmath351 , and @xmath352 days . adopting a magnitude of 5.0 on day 3.6 along our model light curve of @xmath133 wd
, we obtain @xmath353 days and @xmath354 days as listed in table [ fitting_t2_t3_tb_txonoff ] .
we have estimated the ejecta mass of @xmath355 .
the visual ( @xmath71 ) magnitudes are roughly fitted with our model light curve until about day 40 , but are gradually departing from it after that . at this stage ,
the nova probably entered the transition / nebular phase and the deviation comes from the contribution of strong emission lines .
we have also checked the accuracy of wd mass determination that comes from the ambiguity of chemical composition .
we obtain a best - reproducing model of @xmath356 for @xmath278 ( `` ne nova 1 '' ) , and @xmath357 for @xmath280 ( `` ne nova 3 '' ) .
then , we have an approximate relation of @xmath358 for v4743 sgr .
this relation of the wd mass is steeper ( on @xmath33 ) than that for v598 pup and v382 vel .
we may also conclude that the wd mass is @xmath359 for a typical hydrogen content between @xmath284 of v4743 sgr . from the ff method
, we obtain the distance modulus to v4743 sgr , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 14.7 - 0.9 = 13.8,\ ] ] for the @xmath133 wd .
two other attendant models of @xmath110 and @xmath349 wds in figure [ all_mass_v4743_sgr_v1500_cyg_x55z02o10ne03 ] give similar results of @xmath360 for @xmath110 or @xmath361 for @xmath349 wd .
the lc method gives @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v4743~sgr } = -1.1.\ ] ] with @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , we obtain @xmath362_{\rm lc}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm v4743~sgr } = 13.6.\ ] ] kaler - schmidt s law with @xmath354 days and della valle & livio s law with @xmath353 days and @xmath363 yield @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 13.7 ~(13.7),\ ] ] being consistent with the ff and lc results .
then , the distance modulus to v4743 sgr is @xmath365 and the distance is @xmath366 kpc for @xmath367 @xcite .
our estimate is also consistent with the recent results of @xmath368 and @xmath369 kpc by @xcite .
the orbital period of @xmath370 days ( 6.72 hr ) was derived by @xcite from the orbital modulations with a full amplitude of 0.05 mag ( 2003 )
0.15 mag ( 2005 ) .
then , we estimate the epoch when the companion emerges from the nova envelope . from equation ( [ warner_mass_formula ] )
, we have @xmath371 , which corresponds to @xmath372 and @xmath373 for @xmath374 .
it is @xmath375 days at @xmath376 .
v1281 sco ( nova scorpii 2007 no.2 ) was discovered by nakamura on 2007 february 19.86 ut ( jd 2454151.3593 ) at mag about 9.3 @xcite . this object was not detected by nakamura @xcite on february 14.86 ut ( limiting mag about 12.0 ) and also by fujita @xcite on february 18.85 ut ( limiting mag about 11.6 ) , so we here assume that @xmath256 jd 2454148.5 ( 2007 february 17.0 ut ) is the outburst day .
a bright supersoft x - ray phase was reported at day @xmath377 ( 2008 january 24.18 ut ) by @xcite .
to search for the end epoch of supersoft x - ray phase , i.e. , @xmath195 , we use an automatic analyzer in the _ swift _ web page @xcite . the x - ray ( 0.310 kev ) count rates with
the _ swift _ xrt are plotted in figure [ all_mass_v1281_sco_v1500_cyg_x55z02o10ne03 ] together with its optical light curve .
we have checked the early data points of x - ray obtained by automatic analyzer are very similar to the values reported by @xcite .
the epoch of @xmath195 is about 425 days after the outburst , if we regard that the last observation indicates the decay of supersoft x - ray flux .
our best - reproducing light curve of a @xmath378 wd is plotted in figure [ all_mass_v1281_sco_v1500_cyg_x55z02o10ne03 ] for a chemical composition of a typical neon nova , `` ne nova 2 '' , which shows @xmath379 , @xmath380 , and @xmath381 days after the outburst .
we determined the wd mass to be @xmath382 mainly from the supersoft x - ray data . adopting a magnitude of 9.2 on day 3.0 along our model light curve of @xmath378 wd
, we obtain @xmath383 and @xmath384 days as listed in table [ fitting_t2_t3_tb_txonoff ] .
we have estimated the ejecta mass of @xmath385 .
the visual ( @xmath71 ) magnitudes are roughly fitted with our model light curve until about day 80 , but are gradually departing from it after that .
this deviation comes from the contribution of strong emission lines .
we have also examined the accuracy of wd mass determination that comes from the ambiguity of chemical composition .
we obtain a best - reproducing model of @xmath386 for @xmath278 ( `` ne nova 1 '' ) , and @xmath267 for @xmath280 ( `` ne nova 3 '' ) .
then , we have @xmath387 for v1281 sco
. we may conclude that the wd mass is @xmath388 for a typical hydrogen content between @xmath389 0.65 of v1281 sco . from the ff method
, we obtain the distance modulus to v1281 sco , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 18.8 - 1.0 = 17.8,\ ] ] for the @xmath378 wd model .
two other attendant models of @xmath133 and @xmath349 wds in figure [ all_mass_v1281_sco_v1500_cyg_x55z02o10ne03 ] give similar results of @xmath390 for @xmath133 or @xmath391 for @xmath349 wd .
the lc method gives @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v1281~sco } = -5.0.\ ] ] with @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , we obtain @xmath319_{\rm lc}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm v1281~sco } = 17.5.\ ] ] kaler - schmidt s law with @xmath384 days and della valle & livio s law with @xmath383 days and @xmath392 ( at @xmath393 days ) give @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 17.9 ~(17.9),\ ] ] being consistent with the ff and lc results , where the difference between the lc method and the others is about @xmath142 mag but within the ambiguity in this method ( see sections [ prediction_by_t3 ] and [ v598_pup_model ] ) .
then the distance modulus is @xmath394 and the distance to v1281 sco is @xmath395 kpc for @xmath396 @xcite .
v597 pup ( nova puppis 2007 no.1 ) was discovered by pereira on 2007 november 14.23 ut ( jd 2454418.73 ) at mag about 7.0 @xcite .
this object was not detected @xcite on november 6.23 , 7.22 , 8.23 , 10.23 , and 11.22 ut ( limiting mag about 8) , so we assume that @xmath256 jd 2454416.0 ( 2007
november 11.5 ut ) is the outburst day .
a bright supersoft x - ray phase was reported by @xcite .
to search for the end epoch of supersoft x - ray phase , i.e. , @xmath195 , we use an automatic analyzer in the _ swift _ web page @xcite . the x - ray ( 0.3 10 kev ) count rates with
the _ swift _ xrt are plotted in figures [ all_mass_v597_pup_v1500_cyg_x35z02c10o20 ] and [ all_mass_v597_pup_v1500_cyg_x65z02o03ne03 ] as well as the optical light curves .
the x - ray data points show that the turnoff of supersoft x - ray is @xmath397 460 days and the turn - on is probably @xmath398 170 days .
this is a rather rare case in which both the turn - on and turnoff times of supersoft x - ray are specified .
the chemical composition of ejecta is not known , so we first assumed a chemical composition of `` co nova 2 '' and made model light curves as shown in figure [ all_mass_v597_pup_v1500_cyg_x35z02c10o20 ] .
the durations of supersoft x - ray phases are too short to be comparable with the observation .
this is because the duration of supersoft x - ray is shorter for a smaller hydrogen content of @xmath33 .
our reasonably - reproducing model light curve of a @xmath110 wd is plotted in figure [ all_mass_v597_pup_v1500_cyg_x65z02o03ne03 ] for a chemical composition of `` ne nova 3 '' , which shows @xmath399 , @xmath400 , and @xmath401 days after the outburst .
we estimated the wd mass to be @xmath402 mainly from the supersoft x - ray data . adopting a magnitude of 8.4 on day 2.7 as the maximum magnitude along our model light curve of @xmath110 wd
, we obtain @xmath403 and @xmath404 days as listed in table [ fitting_t2_t3_tb_txonoff ] .
we have estimated the ejecta mass of @xmath405 .
the optical light curve shape around the peak deviates largely from our model light curves and is similar to that of v1500 cyg .
it is highly likely that v597 pup is a superbright nova .
after the superbright phase , @xmath71 magnitudes are roughly fitted with our model light curve until day @xmath152 .
we have also checked the accuracy of wd mass determination from the ambiguity of chemical composition . for chemical compositions of `` ne nova 1 '' and `` ne nova 2
, '' however , we could not obtain reasonable fits like those as shown in figure [ all_mass_v597_pup_v1500_cyg_x65z02o03ne03 ] because supersoft x - ray durations of these models are too short to fit with the supersoft x - ray observation like in figure [ all_mass_v597_pup_v1500_cyg_x35z02c10o20 ] .
therefore , we may conclude that the wd mass is @xmath406 and that the composition is relatively hydrogen - rich like @xmath407 . from the ff method
, we obtain the distance modulus to v597 pup , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 17.8 - 0.9 = 16.9,\ ] ] for the @xmath133 wd ( `` ne nova 2 '' ) model .
two other attendant models of @xmath110 and @xmath349 wds ( not shown in the figure for `` ne nova 2 '' models ) , give similar results of @xmath408 for @xmath110 or @xmath409 for @xmath349 wd . from figure [ all_mass_v597_pup_v1500_cyg_x35z02c10o20 ]
, we also have @xmath410 for @xmath349 , @xmath411 for @xmath119 , or @xmath412 for @xmath62 wd with `` co nova 2 . ''
the lc method gives @xmath169_{\rm v1500~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v597~pup } = -4.2.\ ] ] with @xmath182 from equation ( [ distance_modulus_v1500_cyg ] ) for v1500 cyg , we obtain @xmath319_{\rm lc}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm v597~pup } = 16.7.\ ] ] kaler - schmidt s law with @xmath404 days and della valle & livio s law with @xmath403 days , and both with @xmath413 ( at @xmath414 days ) yield @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 17.1 ~(17.1),\ ] ] being consistent with the ff and lc results
. then we have @xmath415 and the distance to v597 pup , @xmath416 kpc , for @xmath417 @xcite .
the orbital period of @xmath418 days ( 2.6687 hr ) was derived by @xcite from the orbital modulations with a full amplitude of 0.2 mag ( 2008 ) 0.6 mag ( 2009 ) .
then , we estimate the epoch when the companion emerges from the nova envelope . from equation ( [ warner_mass_formula ] )
, we have @xmath419 , which corresponds to @xmath420 and @xmath421 for @xmath422 .
it is @xmath423 days at @xmath376 .
@xcite reported that a repetitive hump with an amplitude of @xmath424 mag is seen but no obvious eclipse features in 2008 ( @xmath425 days after outburst ) .
their fourier analysis of light curves provide an orbital period of 2.67 hr .
after that , their 2009 observation revealed broad eclipses with @xmath426 mag deep ( @xmath427 days after outburst ) .
they interpreted that the drop in brightness ( 2.5 mag ) between these two observations in 2008 and 2009 was due largely or entirely to the dispersal and cooling of nova ejecta far from the central binary and the eclipse and their related humps could well have been present even in the earlier 2008 light curves , that is , there are some features resembling such diluted structures .
this interpretation is consistent with the emergence of companion on @xmath342 ( @xmath428 ) or @xmath429 days ( @xmath430 ) after outburst .
moreover , the wd photosphere is larger than that of the companion , i.e. , @xmath431 , during the optically thick wind phase , at least , until 130 days after outburst . after the winds
stop on day @xmath346 , the wd photosphere rapidly shrinks toward its original size , that is , it becomes a half by one month ( see , e.g. , * ? ? ?
v1494 aql ( nova aquilae 1999 no.2 ) was discovered by pereira on 1999 december 1.785 ut ( jd 2451514.285 ) at mag about 6.0 @xcite .
this object was not detected by liller @xcite on november 25.035 ut ( limiting mag about 10.5 ) , so we assume that @xmath256 jd 2451513.0 ( 1999
november 30.5 ut ) is the outburst day .
the star reached its maximum of @xmath432 on 1999 december 3.4 ut ( jd 2451515.9 ) about 3 days after the outburst .
then it declined by 2 mag in 6.6 days and by 3 mag in 16 days ( e.g. , * ? ? ?
the transition oscillation began on day @xmath343 and ended on day @xmath258 , during which the star exhibited quasi - periodic oscillations with periods of 7 to 22 days and with an amplitude reaching 1.5 mag .
we plot the optical and supersoft x - ray light curves in figure [ light_curve_v1494_aql_v1974_cyg_x55z02o10ne03 ] .
a bright supersoft x - ray phase was detected with _
chandra _ on day 250 ( 2000 august 6 ut ) , day 302 ( 2000 september 28 ut ) , and day 305 ( 2000 october 1 ut ) @xcite .
@xcite reported that the supersoft x - ray flux had already declined by a factor of 40 on day @xmath433 ( 2001 november 28 ut ) .
then the supersoft x - ray phase started at least on day @xmath434 and ended before day @xmath435 .
@xcite analyzed their spectra of v1494 aql and estimated the helium / hydrogen ratio to be he / h@xmath436 by number ratio , which corresponds to @xmath437 .
this helium / hydrogen ratio is close to our chemical composition of `` ne nova 2 , '' in which @xmath438 .
@xcite derived he / h@xmath439 , n / h@xmath440 ( or 0.0015 ) , o / h@xmath441 , and fe / h@xmath442 by number ratio .
this corresponds to @xmath443 , @xmath444 , @xmath445 , and @xmath446 , ( or to @xmath447 , @xmath448 , @xmath449 , and @xmath446 ) .
we do not know which is better , so that we adopt an arithmetic average of these two , i.e. , @xmath450 , @xmath451 , @xmath452 , and @xmath446 , which is close to `` ne nova 2 . ''
therefore , we assume `` ne nova 2''as the envelope chemical composition of v1494 aql . our best - reproducing light curve is a @xmath453 wd model among 1.04 , 1.06 , and 1.08 @xmath80 wds mainly from the supersoft x - ray data ( figure [ light_curve_v1494_aql_v1974_cyg_x55z02o10ne03 ] ) , i.e. , @xmath454 . for the @xmath453 wd model ,
we derive @xmath455 , @xmath456 , and @xmath457 days . adopting a magnitude of @xmath252 on day 2.7 as the optical maximum along our model light curve of @xmath453 wd
, we obtain @xmath458 days and @xmath459 days ( table [ fitting_t2_t3_tb_txonoff ] ) .
the visual ( @xmath71 ) magnitudes are not well fitted with our model light curve , from day 10 to day 150 , mainly because the light curve has wavy structures and seems to be strongly contaminated by emission lines .
our model light curve almost follows the bottom line of @xmath71 magnitudes during the transition oscillations , which is supported by figure [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ] for gk per ( see also figure 2 of * ? ? ?
the nova entered a nebular phase @xmath152 days after the outburst and the visual magnitudes deviated much from our model light curve due mainly to strong emission lines such as [ ] ( e.g. , * ? ? ? * ) .
we have also estimated the ejecta mass to be @xmath460 ( table [ physical_properties_of_novae ] ) , which is lost by the optically thick wind ( see , e.g. , * ? ? ?
* ; * ? ? ?
* for more detail ) .
@xcite derived the ejecta mass of @xmath461 , being consistent with our model estimation if we take the distance of @xmath462 kpc as derived below .
based on submillimeter- and centimeter - band fluxes , a single merlin image , and optical spectroscopy , @xcite concluded that optical spectroscopy indicates continued mass ejection for over 195 days , with the material becoming optically thin sometime between 195 and 285 days after the outburst .
this is consistent with @xmath463 days of our @xmath453 wd model .
@xcite reported that v1494 aql had entered into a strong coronal phase at least 230 days after the outburst .
this is closely related to the epoch when the optically thick winds stop , because the photosphere rapidly shrinks and the photospheric temperature quickly increases to @xmath464 k or more .
we have checked the dependency of the wd mass determination on the hydrogen content of the wd envelope .
@xcite reported a different result of helium / hydrogen ratio of the ejecta from 4686/h@xmath465 ratio , that is , he / h@xmath466 by number ratio , which corresponds to @xmath467 and is close to our `` co nova 2 '' ( see table [ chemical_composition ] ) . when we adopt a chemical composition of `` co nova 2 , '' we obtain a best - reproducing model of @xmath468 wd .
the other physical parameters are also listed in tables [ fitting_t2_t3_tb_txonoff ] and [ physical_properties_of_novae ] . on the other hand ,
if we assume a chemical composition of `` ne nova 3 , '' we obtain a best model of @xmath378 wd .
finally we have @xmath469 for v1494 aql . here
, we use @xmath470 , @xmath471 , and @xmath472 .
then , we may conclude that the wd mass is @xmath473 for a typical hydrogen content between @xmath389 0.65 . if the hydrogen content is @xmath474 , then the wd has a mass of @xmath475 and is probably a onemg wd . on the other hand ,
when the hydrogen content is as low as @xmath476 , then the wd is less massive ( @xmath477 ) and probably a co wd . from the ff method
, we obtain the distance modulus to v1494 aql , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 14.9 - 1.5 = 13.4,\ ] ] for the @xmath453 wd ( `` ne nova 2 '' ) model .
two other attendant models of @xmath478 and @xmath479 wds give similar results of @xmath480 for @xmath478 or @xmath481 for @xmath479 wd .
the lc method gives @xmath169_{\rm v1974~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v1494~aql } = -1.3.\ ] ] with @xmath482 from equation ( [ distance_modulus_v1974_cyg ] ) for v1974 cyg , we obtain @xmath319_{\rm lc}= \left[5
\log(d/10 ) + a_v \right]_{\rm v1494~aql } = 13.6.\ ] ] kaler - schmidt s law with @xmath459 days and della valle & livio s law with @xmath458 days and @xmath483 ( at @xmath414 days ) yield @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd}= 13.4 ~(13.6),\ ] ] being consistent with the ff and lc results .
then , we obtain the distance modulus of @xmath484 and the distance of @xmath485 kpc to v1494 aql for @xmath486 @xcite .
this value is a bit larger than the distance of @xmath487 kpc estimated by @xcite from the mmrd relation .
the orbital period of @xmath488 days ( 3.23 hr ) was derived by @xcite from the orbital modulations with an eclipse depth of @xmath489 mag ( see also * ? ? ?
* ; * ? ? ?
then , we estimate the epoch when the companion emerges from the nova envelope . from equation ( [ warner_mass_formula ] ) , we have @xmath490 , which corresponds to @xmath491 and @xmath492 for @xmath493 . these give @xmath494 days at @xmath376 .
various physical parameters are summarized in tables [ fitting_t2_t3_tb_txonoff ] and [ physical_properties_of_novae ] .
@xcite and @xcite reported that higher ionization lines became prominent about 65 days after the optical maximum ( i.e. , day @xmath154 after the outburst ) . at that time
, the photospheric temperature of our @xmath453 wd model has already increased to @xmath495 k and the higher ionization lines mentioned above are possibly related to this high photospheric temperatures .
another possibility is high temperature optically thin plasma that emits hard x - ray mentioned below , because high temperature plasma that emits hard x - ray emerges around @xmath496 days .
hard x - ray was detected with _
long before the supersoft x - ray appeared .
@xcite reported their six _ chandra _ x - ray spectra .
the first three observations were taken with acis - i on 134 , 187 , and 248 days after outburst .
the count rates are 1.0 , 0.69 and 0.53 counts s@xmath261 , respectively .
( we do not plot these point in figure [ light_curve_v1494_aql_v1974_cyg_x55z02o10ne03 ] because the detector acis - i is different from that of supersoft x - ray observation , hrc - s . )
they found no significant variations in any of these observations .
the spectra are dominated by emission lines originating from n , o , and ne .
they fit isothermal apec models with the spectra and found best fits with @xmath497 kev and @xmath498 @xmath499 ( @xmath500 by * ? ? ?
they reported that , in all cases the elemental abundances of o and n had to be significantly increased .
they further reported that on day 248 a bright soft component appeared in addition to the fading emission lines .
when the harder x - ray declined , that is , when the optically thick wind stopped , the soft component of x - ray emerged .
this is very consistent with our collision model between the wind and the companion as an origin of high temperature plasma that emits hard x - ray . in our model ,
the hard component of x - rays emerges around @xmath496 days and attains its peak at @xmath501 days and then declines toward @xmath502 days .
v2467 cyg ( nova cygni 2007 ) was discovered by tago on 2007 march 15.787 ut ( jd 2454175.287 ) at 7.4 mag @xcite .
tago also reported that the nova was fainter than 12th magnitude three days ago ( jd 2454172.296 ) . here
we assume the outburst day of @xmath256 jd 2454172.0 ( 2007 march 12.5 ut ) . in figure [ all_mass_v2467_cyg_v1974_cyg_x55z02o10ne03 ]
, we plot optical and near infrared magnitudes taken from iau circular 8821 , @xcite , and aavso .
the nova reached its maximum of @xmath503 at @xmath504 jd 2454175.2 ( march 16.7 ut ) . except for some early , low amplitude oscillations ,
the light curve declined monotonically until day @xmath341 .
then v2467 cyg started its transition phase until day @xmath166 in the form of a series of several oscillations with a quasi - period of 2022 days and a 0.51.1 mag amplitude .
@xcite obtained two _ swift _ observations .
the first observation was taken on 2008 june 15.17 , about 460 days after the outburst , and obtained an x - ray count rate of 0.017 counts s@xmath261 .
about 100 days later ( 2008 september 23.3 ut @xmath505 566 days after the outburst ) , their second observation yields a count rate of 0.048 counts s@xmath261 .
the both x - ray spectra are soft with a blackbody temperature of @xmath506 ev and @xmath507 ev , respectively .
these two count rates are plotted in figure [ all_mass_v2467_cyg_v1974_cyg_x55z02o10ne03 ] .
to search for the end epoch of supersoft x - ray phase , i.e. , @xmath195 , we use an automatic analyzer in the _ swift _ web page @xcite .
the three x - ray ( 0.3 - 10 kev ) data points with the _ swift _ xrt are plotted in figure [ all_mass_v2467_cyg_v1974_cyg_x55z02o10ne03 ] , early two of which are the same data as those by @xcite .
we have checked that the early two data points of the automatic analyzer are very similar to those obtained by @xcite .
the epoch of @xmath195 is around 760 days after the outburst . our free - free model
light curves are plotted in figure [ all_mass_v2467_cyg_v1974_cyg_x55z02o10ne03 ] for a chemical composition of `` ne nova 2 . ''
the bottom line of @xmath179 and @xmath180 magnitudes are well fitted with our model light curve until about day @xmath152 .
on the other hand , @xmath71 magnitude is finally departing from the @xmath179 and @xmath180 magnitudes from day @xmath166 due to strong emission line contributions .
the best - reproducing wd mass is @xmath479 among 1.08 , 1.04 , and @xmath62 . for the @xmath508 wd model ,
we obtain @xmath509 , @xmath510 , and @xmath511 days .
we also obtain intrinsic @xmath512 and @xmath513 days along our model light curve . here
we assume that the model light curve reached its maximum ( 7.9 mag ) at @xmath514 days after the outburst ( at the first observational point ) .
@xcite reported the possible orbital period of v2467 cyg , @xmath515 days ( 3.82 hr ) , from their unfiltered , @xmath179 , and @xmath180 band ccd photometry .
similar periods of 3.57 hr and 3.85 hr had been already observed in 2007 , several weeks after maximum , by @xcite .
we estimate the epoch when the companion emerges from the nova envelope . with the donor mass of @xmath516 estimated from equation ( [ warner_mass_formula ] ) ,
we obtain @xmath517 and @xmath518 for @xmath519 , giving @xmath520 days at @xmath376 .
we have checked the accuracy of wd mass determination from the ambiguity of chemical composition .
we obtain a best - reproducing model of @xmath521 for @xmath278 ( `` co nova 2 '' ) , and @xmath522 for @xmath280 ( `` ne nova 3 '' ) .
then , we have @xmath523 for v2467 cyg .
thus , we may conclude that the wd mass is @xmath524 for a typical hydrogen content between @xmath389 0.65 . if the hydrogen content is @xmath474 , then the wd has a mass of @xmath525 and is probably a onemg wd . on the other hand ,
when the hydrogen content is as low as @xmath476 , then the wd is less massive ( @xmath526 ) and probably a co wd . from the ff method
, we obtain the distance modulus to v2467 cyg , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 18.0 - 1.7 = 16.3,\ ] ] for the @xmath479 wd ( `` ne nova 2 '' ) model .
two other attendant models of @xmath478 and @xmath62 wds give similar results of @xmath527 for @xmath478 or @xmath528 for @xmath62 wd .
the lc method gives @xmath169_{\rm v1974~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v2467~cyg } = -4.1.\ ] ] with @xmath482 from equation ( [ distance_modulus_v1974_cyg ] ) for v1974 cyg , we have @xmath319_{\rm lc } = \left[5 \log(d/10 ) + a_v \right]_{\rm v2467~cyg } = 16.4.\ ] ] kaler - schmidt s law with @xmath529 days and della valle & livio s law with @xmath530 days and @xmath531 at maximum ( at @xmath532 days ) give @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 16.2 ~(16.5 ) , \label{v1668_cyg_v2467_cyg}\ ] ] being consistent with the ff and lc results .
then , we obtain the distance modulus of @xmath533 and the distance of @xmath485 kpc to v2467 cyg for @xmath534 @xcite .
v5116 sgr ( nova sagittarii 2005 no.2 ) was discovered by liller on july 4.049 ut ( jd 2453555.549 ) at mag about 8.0 @xcite and became a supersoft x - ray source @xcite .
we have already published analysis of optical light curves for this nova @xcite before the supersoft x - ray was detected by @xcite .
this object was not detected on june 12 ( limiting mag about 11.0 ) , so @xcite assumed that @xmath256jd 2453552.0 ( july 1.5 ut ) is the outburst day .
their best - reproducing model is a @xmath535 wd for the chemical composition of `` co nova 2 . ''
adopting 8.0 mag on jd 2453555.549 observed by liller @xcite as the maximum visual magnitude for , hachisu & kato derived @xmath536 days and @xmath537 days . from their best light curve of @xmath535 wd , hachisu & kato predicted that the supersoft x - ray phase starts on @xmath538 days and ends on @xmath539 days after the outburst .
we plot the supersoft x - ray light curves as well as optical in figure [ all_mass_v5116_sgr_v1974_cyg_x65z02o03ne03 ] .
a bright supersoft x - ray phase was first detected with _
xmm newton _
@xcite on day 674 ( 2007 march 5 ut ) .
@xcite reported that the x - ray light curve shows abrupt decreases and increases of the flux by a factor of @xmath540 , which is consistent with a periodicity of 2.97 hr , the orbital period suggested by @xcite , and speculated that the x - ray light curve may result from a partial coverage by an asymmetric accretion disk in a high - inclination system .
@xcite also reported the supersoft x - ray detection with _
swift _ on day 768 ( 2007 august 7.742 ut ) , and @xcite observed a similar supersoft x - ray phase on day 789 ( 2007 august 28 ut ) .
to search for the end epoch of supersoft x - ray phase , i.e. , @xmath195 , we use the automatic analyzer in the _ swift _ web page @xcite . the x - ray ( 0.3 - 10 kev ) count rate with the _ swift _ xrt
are plotted in figure [ all_mass_v5116_sgr_v1974_cyg_x65z02o03ne03 ] .
we have checked that the early several data points of the automatic analyzer are very similar to those obtained by @xcite .
the epoch of @xmath195 is somewhere between 820 and 1070 days after the outburst , being roughly consistent with the prediction by @xcite .
@xcite found , from their x - ray spectrum fit , that oxygen / neon - rich wd atmosphere models provide a better fit than carbon / oxygen - rich wd atmosphere models .
therefore , we adopt here the chemical composition of `` ne nova 3 '' and recalculated model light curves .
our new best - reproducing model is a @xmath541 wd ( figure [ all_mass_v5116_sgr_v1974_cyg_x65z02o03ne03 ] ) , from which we obtain @xmath542 , @xmath543 , @xmath544 days as listed in table [ fitting_t2_t3_tb_txonoff ] .
the estimated wd mass of @xmath541 is very close or nearly equal to the upper limit mass for co wds born in a binary , i.e. , @xmath271 ( e.g. , * ? ? ?
* ) , being consistent with the suggestion that the wd of v5116 sgr is an onemg white dwarf . adopting a magnitude of @xmath545 at maximum ( @xmath546 3.6 days ) along our model light curve of the @xmath541 wd , we obtain `` intrinsic '' @xmath547 and @xmath548 days as listed in table [ fitting_t2_t3_tb_txonoff ] .
the @xmath71 magnitudes are roughly fitted with our model light curve at least until day @xmath152 .
we have checked the accuracy of wd mass determination .
we obtain a best - reproducing model of @xmath549 for @xmath550 ( `` ne nova 2 '' ) , and @xmath149 for @xmath278 ( `` co nova 2 '' ) .
then , we have @xmath551 for v5116 cyg . only when the hydrogen content is @xmath407 , the wd has a mass as large as @xmath541 , which should be consistent with an onemg wd . on the other hand ,
when the hydrogen content is as low as @xmath4760.55 , then the wd is less massive ( 0.85@xmath62 ) and probably a co wd . from the ff method
, we obtain the distance modulus to v5116 sgr , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 18.2 - 2.0 = 16.2,\ ] ] for the @xmath62 wd ( `` ne nova 2 '' ) model .
two other attendant models of @xmath119 and @xmath61 wds ( `` ne nova 2 '' ) give similar results of @xmath552 for @xmath119 or @xmath553 for @xmath61 wd .
the lc method gives @xmath554_{\rm v1974~cyg } - \left[5 \log(d ) + a_v \right]_{\rm v5116~sgr } = -3.9.\ ] ] with @xmath482 from equation ( [ distance_modulus_v1974_cyg ] ) for v1974 cyg , we obtain @xmath319_{\rm lc}= \left [ 5 \log(d ) + a_v \right]_{\rm v5116~sgr } = 16.2.\ ] ] kaler - schmidt s law with @xmath548 days and della valle & livio s law with @xmath547 days , and both with @xmath555 at maximum ( @xmath556 days ) yield @xmath364_{\rm mmrd}= \left [ 5 \log(d ) + a_v \right]_{\rm mmrd } = 16.0 ~(16.2),\ ] ] being consistent with the ff and lc results .
then the distance modulus is @xmath557 and the distance to v5116 sgr is @xmath558 kpc for @xmath559 @xcite .
these distance estimates are consistent with @xmath560 kpc discussed by @xcite .
the orbital period of @xmath561 days ( 2.97 hr ) was derived by @xcite from the orbital modulations with a full amplitude of @xmath562 mag .
then , we estimate the epoch when the companion emerges from the nova envelope . from equation ( [ warner_mass_formula ] )
, we obtain @xmath563 , which yields @xmath564 and @xmath565 for @xmath566 and , as a result , @xmath567 days at @xmath376 .
v574 pup ( nova puppis 2004 ) was discovered independently by tago and sakurai on 2004 november 20.672 ut ( jd 2453330.172 ) at mag about 7.6 @xcite .
this object was not detected on november 12 and 16 with limiting mag about 13.6 @xcite , so we assume that @xmath256jd 2453326.0 ( november 16.5 ut ) is the outburst day .
the optical and near infrared light curves are plotted in figure [ all_mass_v574_pup_v1668_cyg_x65z02o03ne03 ] .
although @xcite derived @xmath568 on jd 2453335.6 ( november 26.1 ut ) , this is not the @xmath71 maximum but the second peak as can be seen in figure [ all_mass_v574_pup_v1668_cyg_x65z02o03ne03 ] .
we derive @xmath569 at @xmath504 jd 2453332.0 ( november 22.5 ut ) , 6 days after the outburst based on the data reported in iau circular 8445 @xcite .
after the @xmath71 magnitude drops by about 2 mag at @xmath570 jd 2453333.7 ( november 24.2 ut ) , it increases to 8.0 mag at @xmath570 jd 2453335.6 ( november 26.1 ut ) .
after that , the @xmath71 magnitude declines to about 9.0 mag and stays for 20 days and then declines almost monotonically . in this way
, v574 pup shows an oscillatory feature in the very early phase .
_ swift _ xrt observations of v574 pup showed that the nova became a supersoft x - ray source before 2005 august 17 @xcite , that is , 274 days after the outburst .
@xcite reported that v574 pup was still in a supersoft x - ray phase on 2007 december 7 and 13 , that is , 1121 days after the outburst . to search for the end epoch of supersoft x - ray phase
, we use the automatic analyzer in the _ swift _ web page @xcite . the x - ray ( 0.3 - 10 kev ) count rate with the _ swift _ xrt
are plotted in figure [ all_mass_v574_pup_v1668_cyg_x65z02o03ne03 ] .
we added the last observational point from _ swift _ web page .
the epoch of @xmath195 is somewhere between 1130 and 1320 days after the outburst .
@xcite suggested that the observational properties of v574 pup were intermediate between those typical of explosions on a co or an onemg wd .
their _ spitzer _
observation shows that v574 pup reveals the strong coronal lines of mg and ne ions ( but no at 12.8 microns ) , the behavior of which is suggestive of an explosion on the surface of an onemg wd .
therefore , we have calculated light curves for three different chemical compositions , i.e. , `` ne nova 3 , '' `` ne nova 2 , '' and `` co nova 2 . '' however , only the case of `` ne nova 3 '' provides a reasonable fit to the observation .
our best - reproducing light curve is plotted in figure [ all_mass_v574_pup_v1668_cyg_x65z02o03ne03 ] for the chemical composition of `` ne nova 3 . ''
the wd mass is estimated to be @xmath571 , which suggests a possibility that the wd is an onemg wd , because the estimated wd mass of @xmath119 is marginally consistent with the theoretically obtained upper limit mass for co wds born in binaries , @xmath271 ( e.g. , * ? ? ?
* ) . for the other two chemical compositions ,
i.e. , `` ne nova 2 , '' and `` co nova 2 , '' we do not find any reasonable light curves which simultaneously satisfy both the start and end of supersoft x - ray phase , that is , the duration of supersoft x - ray phase is too short to be compatible with the observation , like as has already shown in figure [ all_mass_v597_pup_v1500_cyg_x35z02c10o20 ] .
we determine @xmath192 and @xmath3 along our model light curve . assuming that our model light curve reached its maximum at the discovery day ( @xmath572 days after the outburst ) , we obtain `` intrinsic '' @xmath573 and @xmath574 days .
the other timescales are obtained to be @xmath575 , @xmath576 , @xmath577 days ( table [ fitting_t2_t3_tb_txonoff ] ) . from the ff method , we obtain the distance modulus to v574 pup , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 17.4 - 2.0 = 15.4,\ ] ] for the @xmath62 wd ( `` ne nova 2 '' ) model , although the models of `` ne nova 2 '' are not shown in the figure .
two other attendant models of @xmath479 and @xmath61 wds ( `` ne nova 2 '' ) give similar results of @xmath578 for @xmath479 or @xmath579 for @xmath61 wd .
the lc method gives @xmath169_{\rm v1974~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v574~pup } = -3.2.\ ] ] with @xmath482 from equation ( [ distance_modulus_v1974_cyg ] ) for v1974 cyg , we have @xmath319_{\rm lc}= \left[5 \log(d/10 ) + a_v \right]_{\rm v574~pup } = 15.5.\ ] ] kaler - schmidt s law with @xmath580 days or della valle & livio s law with @xmath573 days and @xmath581 at maximum ( @xmath582 days ) give @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 15.5 ~(15.7),\ ] ] being consistent with the ff and lc results .
thus the distance modulus is @xmath583 and the distance to v574 pup is @xmath584 kpc for @xmath585 @xcite .
v458 vul ( nova vulpeculae 2007 ) was discovered by abe on 2007 august 8.54 ut ( jd 2454321.04 ) at mag about 9.5 @xcite .
this object was not detected on july 23 , 31 , and august 4 with limiting mag about 11.5 @xcite , so we assume that @xmath256jd 2454317.0 ( august 4.5 ut ) is the outburst day . the light curve is shown in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] .
the @xmath71 magnitudes show a striking multi - peak ( or multi - pulse ) structure in the early phase .
this kind of multi - peak structures were observed in hr del 1967 and v723 cas 1995 , both of which are considered to be a nova on a co wd .
so we assume here that v458 vul was also a nova on a co wd .
@xcite reported the supersoft x - ray detection of v458 vul with the _ swift _ xrt on 2007 october 18.05 ut ( 75 days after the outburst ) .
@xcite , however , concluded , with the _ suzaku _ observations , that the spectrum can be described by a single temperature ( @xmath586 kev ) thin thermal plasma model , suggesting shock - origin of the x - ray .
recently , @xcite reported , with the _ swift _ xrt observations obtained from 2008 june 18 to september 16 ( about @xmath587 days after the outburst ) , that v458 vul have entered a new phase characterized by a highly variable supersoft x - ray component accompanied by partially anti - correlated variations in the ultraviolet .
they also pointed that an earlier report on entry into the supersoft phase @xcite has proven premature .
@xcite reported a summary of the early phase x - ray observation with _
swift _ in which a supersoft x - ray phase started about 400 days after the outburst .
figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] shows the supersoft x - ray fluxes reported in @xcite .
the spectra show hard ( _ star marks _ ) components similar to tsujimoto et al.s until day @xmath348 and then turned into supersoft ( _ open diamonds _ ) .
the chemical composition of v458 vul is not yet reported .
so we here assume `` co nova 4 '' in table [ chemical_composition ] , which is similar to that of hr del @xcite and v723 cas @xcite . our model
light curves are plotted in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] .
this nova has an oscillation phase in the early @xmath71 , @xmath179 , and @xmath180 light curves .
we thus fit our universal decline law with the bottom line of each oscillation as shown in the figure .
the bottom line of @xmath180 magnitudes is nicely fitted with our model light curve until day @xmath152 . on the other hand , the other two , @xmath71 and @xmath179 magnitudes , are gradually departing from the model light curve due to emission line contributions . in the previous subsections , we have determined the wd masses mainly from x - ray light curve fitting because optical and near ir model light curves ( i.e. , free - free emission model light curves ) have a similar shape of light curves and are not sensitive to the wd mass as explained in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20 ] . in the case of v458 vul ,
only the emergence of supersoft x - ray was clearly detected but the supersoft x - ray phase does not end yet .
therefore , we have to estimate the wd mass only from the @xmath197 time .
the supersoft x - ray flux replaced the hard x - ray component and rose up around @xmath588 days as shown in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] .
as shown in v382 vel , we proposed a hypothesis that hard x - ray component is a shock - origin by optically thick winds .
therefore , the substantial flux of hard x - ray indicates that optically thick winds still blow during the hard x - ray phase .
on the other hand , supersoft x - ray may be self - absorbed by wind itself .
therefore , the emergence of supersoft x - ray indicates that the optically thick wind has stopped or substantially weakened ( e.g. , * ? ? ?
our supersoft x - ray model light curves show a gradual rise in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] because our flux is the photospheric flux and does not include the effect of absorption by winds . here
we suppose that supersoft x - ray is absorbed during the optically thick wind phase .
the sudden rise in the supersoft x - ray flux just after the optically thick winds stopped matches our view of dense optically thick winds ( see , e.g. , * ? ? ? * for a sharp increase / decrease supersoft x - ray flux of rx j0513.9@xmath589 ) .
the date of supersoft x - ray emergence is sharply determined to be @xmath590 days for v458 vul .
we have calculated three models of 0.95 , 0.93 , and @xmath535 wds and obtained @xmath591 , 416 , and 456 days , respectively .
then the wd mass is estimated to be @xmath592 for `` co nova 4 '' from the emergence time of supersoft x - ray , i.e. , @xmath593 days . here
we regard that the sharp increase in the supersoft x - ray flux corresponds to the end epoch of optically thick winds at about day 415 . for the best model of @xmath594 among @xmath139 , @xmath595 , and @xmath535 wds
, we obtain @xmath596 days , @xmath597 days , and @xmath598 days ( 2011 january 5.5 ut ) as listed in table [ fitting_t2_t3_tb_txonoff ] .
we have checked the accuracy of wd mass determination especially for cno abundance .
we adopt other chemical composition of `` ne nova 2 , '' same hydrogen content @xmath550 but different cno abundance @xmath599 vs. 0.2 , and we obtain @xmath600 , 430 , and 531 days for @xmath601 , 0.95 , and @xmath535 wds , respectively . here
we adopt the best - reproducing model of @xmath145 among these three wd masses as shown in figure [ all_mass_v458_vul_v1974_cyg_x55z02o10ne03 ] .
this result may suggest that @xmath602 for v458 vul .
the slope of @xmath603 for cno abundance is much smaller than that ( @xmath426 ) for hydrogen content .
therefore , if we regard the usual range of chemical composition as @xmath604 and @xmath605 , the ambiguity in wd mass determination ( @xmath606 ) for cno abundance is much smaller than that ( @xmath607 ) for hydrogen content .
therefore it is likely that @xmath608 for a broad range of the chemical composition @xmath604 ( @xmath609 ) and @xmath605 ( @xmath610 ) .
we have estimated @xmath192 and @xmath3 along our model light curve . placing our model light curve on the bottom line of multi - pulse light curves as shown in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] and assuming that the @xmath71 magnitude attains its maximum of @xmath611 mag at @xmath612 days
, we obtain `` intrinsic '' @xmath613 and @xmath614 days .
the assumed value of our theoretical peak is @xmath615 mag dimmer than that of the observed peak ( 8.0 mag ) . from the ff method
, we obtain the distance modulus to v458 vul , i.e. , @xmath167_{\rm ff } = m_{\rm w } - m_{\rm w } = 19.3 - 2.3 = 17.0,\ ] ] for the @xmath61 wd ( `` ne nova 2 '' ) as shown in figure [ all_mass_v458_vul_v1974_cyg_x55z02o10ne03 ] .
two other attendant models of @xmath62 and @xmath535 wds give similar results of @xmath616 for @xmath62 or @xmath617 for @xmath535 wd . here
we do not use the model light curves of `` co nova 4 , '' because they are not calibrated yet .
the lc method gives @xmath169_{\rm v1974~cyg } - \left[5 \log(d/10 ) + a_v \right]_{\rm v458~vul } = -4.7.\ ] ] with @xmath482 from equation ( [ distance_modulus_v1974_cyg ] ) for v1974 cyg , we have @xmath319_{\rm lc}= \left[5 \log(d/10 ) + a_v \right]_{\rm v458~vul } = 17.0.\ ] ] kaler - schmidt s law with @xmath618 days or della valle & livio s law with @xmath619 days , and both with @xmath620 at maximum ( @xmath621 days ) yield @xmath364_{\rm mmrd}= \left [ 5 \log(d/10 ) + a_v \right]_{\rm mmrd } = 17.1 ~(17.3),\ ] ] being consistent with the ff and lc results .
thus we obtain the distance modulus of @xmath622 and the distance of @xmath623 kpc to v458 vul for @xmath624 @xcite .
this distance is roughly consistent with @xmath625 kpc estimated by @xcite .
finally , we introduce some characteristic features concerning binary nature , that is , the emergence of the companion from the wd photosphere . @xcite reported the orbital period of @xmath626 day with a full amplitude of 0.4 mag .
the orbital light curve shows a saw - tooth shape similar to those in other supersoft x - ray novae , e.g. ci aql and v723 cas .
because equation ( [ warner_mass_formula ] ) is not applicable to such a long orbital period of @xmath627 hr , we adopt a condition that the companion mass is smaller than @xmath628 .
this condition comes from @xmath629 ( @xmath630 ) required for thermally stable mass transfer in a binary ( e.g. , * ? ? ?
if we assume @xmath631 as an upper limit of the secondary mass , then we obtain @xmath632 and @xmath633 .
it is @xmath634 days at @xmath376 .
even if we take a smaller mass of @xmath635 , we obtain similar values of @xmath636 , @xmath637 , giving @xmath638 days at @xmath376 .
@xcite explained the early _ swift _ x - ray data as shock - origin , which started about 70 days after the outburst and peaked at about 140 days and then decreased as can be seen in figure [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] .
the shock may be formed by collision between two ejecta shells ( e.g. , * ? ? ?
* ; * ? ? ?
* ) or by collision between nova winds and the companion star @xcite .
if these x - rays come from the shock between the optically thick wd wind and the companion , the emergence of hard x - rays should be coincident with the emerge of the companion from the wd photosphere .
this is roughly consistent with @xmath638 days . on the other hand ,
the break of free - free emission light curve is caused by a more steeper decrease in the wind mass - loss rate after the break point .
therefore , the drop of x - ray flux at / near 140 days after the outburst is also consistent with our model of @xmath596 days .
the hard x - ray lasted at least until about @xmath639 days , that is , until when the optically thick wind stopped .
it should be noted that , for the orbital period of v458 vul , r. wesson et al .
( private communication ) proposed another orbital period of 98.1 min .
if it is the case , the orbital period of @xmath640 day ( 1.635 hr ) is much shorter than that obtained by @xcite and this period gives @xmath641 from equation ( [ warner_mass_formula ] ) , @xmath642 , @xmath643 , and @xmath644 days ( at @xmath376 ) .
then , the emergence time of the companion star is too late , that is , @xmath645 days , and our model on the origin of hard x - ray may not be viable . in the previous subsections ,
we have fitted our model light curves with those of ten classical novae , in which a supersoft x - ray phase was detected , and estimated various nova parameters .
figure [ supersoft_x_ray_time ] shows a summary of our results , that is , @xmath198 , @xmath197 , and @xmath192 ( from top to bottom ) against @xmath3 for the ten novae , excluding v598 pup but instead including v1974 cyg , which was studied by @xcite .
this figure shows a very good agreement between the data of `` intrinsic '' @xmath646 and equation ( [ t_2-t_3 ] ) .
this is not surprising because equation ( [ t_2-t_3 ] ) can be derived from the universal decline law with @xmath218 as shown in figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] and [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] ( see also * ? ? ?
* ) , and both @xmath192 and @xmath3 are calculated from our model light curve of the universal decline law .
figure [ supersoft_x_ray_time ] also shows reasonable agreement of the observed @xmath197 and @xmath198 against @xmath3 with equations ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) , respectively . theoretically , @xmath197 ( @xmath81 the epoch when optically thick winds stop ) and @xmath198 ( @xmath81 the epoch when hydrogen - shell burning stops ) are uniquely determined if the wd mass and the chemical composition are given
. however , @xmath192 and @xmath3 times are theoretically not unique even if both the wd mass and the chemical composition are given .
these timescales depend further on the initial envelope mass as already explained in section [ properties_of_light_curve ] .
therefore , we must obtain @xmath192 and @xmath3 times from each outburst of various novae . good agreement between our prediction formulae and
the observation on @xmath197 and @xmath198 suggests that our nova models of optically thick winds and free - free model light curves are reasonable for the entire description of nova evolutions and , at the same time , the scatter in the initial nova envelope mass is not so large from their average . in our modeling , we estimated their average relations , @xmath3 vs. @xmath22 or @xmath3 vs. @xmath195 , from the model of v1668 cyg .
such a relatively small scatter of the 10 novae from these relations strongly suggests that we reasonably predict the supersoft x - ray phases of novae from this diagram , i.e. , from equations ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) if we can estimate accurate `` intrinsic '' @xmath192 or @xmath3 time from early optical light curves .
llllcclllll v598 pup 2007#2 & ... & & & & 11.7 & 12.0 & ( ) & 0.27 & @xmath647 & 1 + v382 vel 1999 & ... & 6.6 & 12.4 & 2.6 & 11.4 & 11.5 & 11.6 ( 11.4 ) & 0.62 & @xmath648 & 2 + v4743 sgr 2002#3 & ... & 9.4 & 17 & 5.0 & 13.8 & 13.6 & 13.7 ( 13.7 ) & 0.78 & @xmath649 & 3 + v1281 sco 2007#2 & ... & 9.5 & 16.7 & 9.2 & 17.8 & 17.5 & 17.9 ( 17.9 ) & 2.17 & @xmath650 & 4 + v597 pup 2007#1 & ... & 8.7 & 16.5 & 8.4 & 16.9 & 16.7 & 17.1 ( 17.1 ) & 0.93 & @xmath651 & 5 + v1494 aql 1999#2 & ... & 11.7 & 22.1 & 5.0 & 13.4 & 13.6 & 13.4 ( 13.6 ) & 1.83 & @xmath652 & 6 + v2467 cyg 2007 & ... & 12.0 & 23.0 & 7.9 & 16.3 & 16.4 & 16.2 ( 16.5 ) & 4.65 & @xmath652 & 7 + v5116 sgr 2005#2 & ... & 15.6 & 28.5 & 7.9 & 16.2 & 16.2 & 16.0 ( 16.2 ) & 0.81 & @xmath653 & 8 + v574 pup 2004 & ... & 18.6 & 35.0 & 7.6 & 15.4 & 15.5 & 15.5 ( 15.7 ) & 2.2 & @xmath654 & 8 + v458 vul 2007 & ... & 20.7 & 38.7 & 9.4 & 17.0 & 17.0 & 17.2 ( 17.3 ) & 1.86 & @xmath655 & 9 + + v1974 cyg 1992 & ... & 16 & 26 & 4.2 & & [ 12.3 ] & 12.4 ( 12.5 ) & 1.00 & 1.8 & 10 + v1668 cyg 1978 & ... & 14.4 & 26 & 6.2 & & [ 14.3 ] & 14.4 ( 14.6 ) & 1.24 & 4.1 & 11 + v1500 cyg 1975 & ... & 7.2 & 13 & 3.6 & & [ 12.5 ] & 12.6 ( 12.4 ) & 1.60 & 1.5 & 12 + gk per 1901 & ... & 8.1 & 14.5 & 0.5 & & [ 9.2 ] & 9.3 ( 9.3 ) & 0.93 & 0.455 & 13
our main conclusions consist of three parts : the first part is for basic properties of our model light curves which may be summarized as follows : * we model optical and infrared light curves with free - free emission , and uv 1455 and supersoft x - ray light curves with blackbody emission , based on the optically thick wind model , as done in our previous paper @xcite .
the model parameters governing the light curves are the wd mass ( @xmath32 ) , chemical composition of the wd envelope ( @xmath33 , @xmath34 , @xmath37 , @xmath35 , and @xmath36 ) , and the initial envelope mass ( @xmath199 ) . * our free - free emission model light curves reasonably well
reproduce optical and near ir light curves of novae .
the decay timescale of our free - free emission model light curves depends strongly on the wd mass but weakly on the chemical composition . so the nova speed class is mainly determined by the wd mass . *
the peak brightness of our model light curve depends also on the initial envelope mass ( @xmath199 ) even if we fix the wd mass and the chemical composition of the envelope .
this indicates that nova speed class , i.e. , @xmath192 and @xmath3 times , depends also on the initial envelope mass , in other words , the mass accretion rate to the wd . * our uv 1455 model light curves for different wd masses are homologous ,
that is , they are almost overlapped to each other if we stretch them in the direction of time ( @xmath29 ) by a factor of @xmath90 and properly normalize the flux . all the free - free emission model light curves for different wd masses are also homologous and overlapped with each other if we rescale them as @xmath656 in the time and @xmath657 in the flux ( see figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law_logt ] and [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_new_model ] ) , that is , they are overlapped to each other in the @xmath658@xmath79 plane , where @xmath659 const .
the more massive the wd , the smaller the timescaling ( or stretching ) factor of @xmath66 .
we call this `` a universal decline law . ''
* using the universal decline law , the apparent magnitudes of various wd mass models are restored by @xmath660 and their absolute magnitudes are calculated from @xmath661 .
therefore , we can determine the absolute magnitudes of free - free emission model light curves by calibrating their absolute magnitudes with the known distance moduli @xmath143 of v1668 cyg for co novae and v1974 cyg for neon novae . *
since the absolute magnitudes of our free - free emission model light curves are now calibrated ( tables [ light_curves_of_novae_co ] and [ light_curves_of_novae_ne ] ) , we can estimate the distance moduli of individual novae by directly fitting them with observation .
we show that the distance moduli thus obtained for v1500 cyg are consistent with those calculated from the nebular expansion parallax method .
this is a new method for obtaining the distance modulus to a nova . *
it has been empirically proposed that the absolute @xmath71 magnitude 15 days after optical maximum , @xmath111 , is almost the same among various novae .
we show that there is only a small scatter of 0.10.3 mag for the absolute brightnesses of our model light curves 15 days after maximum ( see figures [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_real_scale_logt_no2 ] and [ mass_v_uv_x_v1974_cyg_x55z02o10ne03_real_scale_model ] ) .
this results theoretically support the above empirical relation on @xmath111 .
* we found that , when the timescales of two nova light curves are similar , the absolute brightnesses of the two novae are almost the same during the period in which the universal decline law is applied .
therefore , we are also able to estimate the distance modulus of a nova by comparing it with the other nova the distance of which is known and the timescale of which is similar . we also show that the distance moduli thus obtained for gk per are consistent with those calculated from the nebular expansion parallax method .
this is another new method for obtaining the distance modulus to a nova . *
our free - free model light curve reasonably follows the observed light curve when free - free emission dominates the continuum flux in novae .
even when the nova has a transition oscillation or a characteristic secondary peak , our free - free emission model light curve can be fitted with the bottom line of observed nova light curves ( see figures [ all_mass_gk_per_v1500_cyg_x55z02o10ne03 ] , [ all_mass_v1493_aql_v1500_cyg_x55z02o10ne03 ] , and [ all_mass_v458_vul_v1974_cyg_x55z02c10o10 ] ) .
the second part contains timescales of novae , our prediction formulae of supersoft x - ray on and off times , and our theoretical mmrd relations , which are derived from the universal decline law : * we derive several relations between the characteristic timescales of classical novae , i.e. , @xmath193 , @xmath22 , and @xmath195 , which depend mainly on the wd mass and weakly on the chemical composition but almost independent of the initial envelope mass . *
on the other hand , @xmath192 and @xmath3 depend sharply on the initial envelope mass . adopting a typical ( average ) case of the initial envelope mass ,
we obtain `` average '' relations between @xmath3 ( or @xmath192 ) time and the three timescales of @xmath193 , @xmath22 , and @xmath195 .
these proposed relations are reasonably fitted with published data of novae .
we propose the relation between the `` intrinsic '' @xmath3 time and @xmath22 , i.e. , equation ( [ t_3-t_wind ] ) , as a prediction formula for the start of a supersoft x - ray phase , and the relation between the intrinsic @xmath3 and @xmath195 , i.e. , equation ( [ t_3-t_h_burning ] ) , as a prediction formula for the end of it . here
we propose the `` intrinsic '' @xmath3 time which is calculated along our free - free emission model light curve fitted with the observation .
* we obtain a relation between the maximum magnitude , @xmath662 , and the rate of decline , @xmath663 , through the stretching factor @xmath66 of the free - free emission model light curves , i.e. , equation ( [ theoretical_apparent_mmrd_relation ] ) .
a more massive wd corresponds to a smaller stretching factor of @xmath66 while the initial envelope mass determines a set of @xmath664 .
calibrating the distance moduli @xmath143 of v1668 cyg and v1974 cyg , we derive usual mmrd relations between the absolute maximum magnitude @xmath665 and @xmath239 by changing @xmath66 for a given set of @xmath664 , i.e. , equation ( [ theoretical_mmrd_relation ] ) . *
our mmrd relations theoretically confirm that the main parameter is the wd mass , @xmath32 . a more massive wd with a smaller @xmath66 corresponds to a faster nova , i.e. , a shorter @xmath3 time .
however , there is a second parameter , i.e. , the initial envelope mass @xmath199 , corresponding to @xmath664 .
the larger the envelope mass , the brighter the maximum magnitude , and the shorter the rate of decline .
the main parameter @xmath32 produces a main trend of mmrd relation and the second parameter @xmath199 can reasonably explain the scatter around the main mmrd relation . * our mmrd relations are also in good agreement with kaler - schmidt s law @xcite and della valle & livio s ( 1995 ) law at their central lines ( values ) .
the last part concerns light - curve fitting of ten classical novae in which a supersoft x - ray phase was detected : * the universal decline law reasonably reproduces the observed light curves of ten classical novae , in which a supersoft x - ray phase was detected .
the models reproducing simultaneously the optical and supersoft x - ray observations are wds with @xmath5 ( v598 pup ) , @xmath6 ( v382 vel ) , @xmath7 ( v4743 sgr ) , @xmath8 ( v1281 sco ) , @xmath9 ( v597 pup ) , @xmath10 ( v1494 aql ) , @xmath11 ( v2467 cyg ) , @xmath12 ( v5116 sgr ) , @xmath13 ( v574 pup ) , @xmath14 ( v458 vul ) .
the first nine wds are probably onemg wds ( one novae ) while the last one is probably a co wd ( co nova ) .
our light curve fittings also suggest various physical values of novae , such as the envelope mass of @xmath199 at optical maximum , the ejecta mass of @xmath277 , etc . as tabulated in table [ physical_properties_of_novae ] . * the newly proposed relationships of equations ( [ t_3-t_wind ] ) and ( [ t_3-t_h_burning ] ) are consistent with the emergence / decay epoch of the supersoft x - ray phase of these novae including v1974 cyg .
* we estimate distances to the ten classical novae by using our two new methods and show that they are consistent with those derived from the mmrd relation based on the `` intrinsic '' @xmath3 time as tabulated in table [ distance_moduli_of_novae ] . *
we propose a mechanism of early hard x - ray emission originated from the shock between the optically thick winds and the companion star .
our model predicts that the hard x - rays emerge when the companion appears from the white dwarf photosphere and decay when the optically thick winds stop .
our estimates of emerging - time @xmath327 of the companion from the nova photosphere are consistent with the observed hard x - ray appearance in v382 vel , v1494 aql , and v458 vul ; which may support our idea of hard x - ray emission .
we thank aavso , afoev , vsnet , and vsolj for nova photometric data .
we are also grateful to the anonymous referee for useful comments that helped to improve the paper .
this work made use of data supplied by the uk swift science data centre at the university of leicester .
this research has been supported in part by the grant - in - aid for scientific research ( 20540227 ) of the japan society for the promotion of science .
in this appendix , we present a formulation of scaling law for free - free emission model light curves and derive equation ( [ simple_final_scaling_flux ] ) . we determine the flux of our free - free light curves against that of @xmath61 wd ( @xmath81 v1668 cyg ) .
now we explicitly express the proportionality constant in equation ( [ wind - free - free - emission ] ) as @xmath666^{\{m_{\rm wd}\}}_{(t)}. \label{explicit - wind - free - free - emission}\ ] ] in free - free emission , @xmath667 is independent of frequency @xmath23 , then @xmath668 is also independent of frequency because each term in the square bracket of the right - hand - side is also independent of frequency .
the proportionality constant @xmath668 probably depends not only on the wd mass but also weakly on the chemical composition and other physical conditions . however , we here assume that @xmath668 depends only on @xmath32 when we fix the chemical composition .
then , we obtained @xmath75 from equation ( [ time - derivation - flux ] ) after time - stretching ( @xmath73 and @xmath74 ) .
since the free - free flux is independent of the frequency @xmath23 , i.e. , @xmath669 , this simply means @xmath670 now , we define the apparent @xmath71 magnitudes as @xmath671 and , for 0.95 @xmath80 wd ( @xmath81 v1668 cyg ) with @xmath672 , we have @xmath673 where we introduce a constant @xmath674 in order to fit @xmath675 of @xmath61 wd model with the @xmath71 observation of v1668 cyg , and @xmath674 is the same among various wd masses .
using equation ( [ scaled - wind - free - free - emission ] ) , we have @xmath676 by the way , we have already defined the apparent @xmath71 magnitudes of our free - free emission model light curves as @xmath677^{\{m_{\rm wd}\}}_{(t / f_{\rm s } ) } + k_v , \label{scaling_flux}\ ] ] in equation ( [ simple_scaling_flux ] ) . here
@xmath83 is a constant defined to fit @xmath79 of @xmath61 wd model with the @xmath71 observation of v1668 cyg , and common among all the wd mass models ( from @xmath55 to 1.2 @xmath80 by 0.05 @xmath80 step ) for a chemical composition of `` co nova 2 . ''
we saw that all the light curves are almost overlapped to each other in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
this simply means that @xmath678^{\{m_{\rm wd}\}}_{(t
/ f_{\rm s } ) } \approx \left [ { { \dot m_{\rm wind}^2 } \over { v_{\rm ph}^2 r_{\rm ph } } } \right]^{\{0.95~m_\sun\}}_{(t)}. \label{overlapping - flux - free - free - emission}\ ] ] substituting equation ( [ explicit - wind - free - free - emission ] ) into equation ( [ scaled - wind - free - free - emission ] ) and using equation ( [ overlapping - flux - free - free - emission ] ) , we have @xmath679 from equation ( [ overlapping - flux - free - free - emission ] ) together with equations ( [ 095_apparent_flux_def ] ) and ( [ scaling_flux ] ) , we derive @xmath680 therefore , we may regard that @xmath681 , or in other words , @xmath682 if we define @xmath83 as @xmath683 thus , we derive equation ( [ simple_final_scaling_flux ] ) .
therefore , we plotted all the brightnesses of novae in the @xmath73 vs. @xmath684 diagram as shown in figure [ v1668_cyg_vy_jhk_mag_x35z02c10o20_model_scaling_law ] .
it should be noted that the uv 1455 and supersoft x - ray model light curves follow @xmath685 after stretching but do not satisfy @xmath686 , so that these fluxes do not obey equations ( [ scaled - wind - free - free - emission ] ) and ( [ final_scaling_flux ] ) , because the blackbody emissivity depends on the frequency @xmath23 and , as a result , the spectrum changes after time - stretching ( @xmath74 ) .
therefore , we have to normalize the uv fluxes by different factors to fit the peaks with the observational peak of v1668 cyg , as already mentioned in section [ properties_of_light_curve ] .
the photospheric radius of a typical nova during a supersoft x - ray phase is usually much larger than the chandrasekhar radius .
it is about ten times larger ( @xmath687 cm ) than the chandrasekhar radius at the beginning of a supersoft x - ray phase and quickly shrinks with the photospheric temperature being increasing .
therefore , we here present a simple formula on the radius of wd photosphere during a supersoft x - ray phase , which may be used for estimating the x - ray emission area . the photospheric radius of the wd envelope in the supersoft x - ray phase is plotted in figure [ st_mariko ] against the photospheric temperature for various wd masses with the chemical composition of `` co nova 2 '' and `` ne nova 2 . ''
the left edge of each line corresponds to the epoch when the wind stops ( at @xmath688 when the supersoft x - ray emerges ) . as the photosphere shrinks , the photospheric temperature rises .
the rightmost point of each line denotes the epoch when hydrogen nuclear burning stops .
we see that the photospheric radius @xmath27 is almost inversely proportional to a square of the photospheric temperature @xmath17 , i.e. , @xmath689 , because the total luminosity is almost constant , @xmath690 const .
, in the supersoft x - ray phase except the very later phase
. the radius of a typical fast classical nova ( @xmath691 wd with a composition of `` co nova 2 '' ) can be approximated by the black solid line as @xmath692 for @xmath693 ( k ) @xmath694 .
the dashed lines denote the upper and lower bound values of @xmath695 which covers the most of the wd mass representative for classical novae , i.e. , from @xmath696 with `` co nova 2 '' to @xmath697 with `` ne nova 2 . ''
the slow novae are not covered in the region , i.e. , @xmath698 with `` co novae 2 '' has smaller photospheric radius as shown in figure [ st_mariko ] .
this figure also shows the wd radius , i.e. , the bottom of a hydrogen burning zone : the chandrasekhar radius for @xmath699 and values taken from @xcite for @xmath700 .
arkhipova , v. p. , burlak , m. a. , & esipov , v. f. 2002 , astronomy letters , 28 , 100 augusto , a. , & diaz , m. p. 2003
, , 125 , 3349 balman , . , retter , a. , & bos , m. 2006 , , 131 , 2628 barsukova , e. a. , & goranskii , v. p. 2003
, astronomy letters , 29 , 195 bonifacio , p. , selvelli , p. l. , & caffau , e. 2000 , , 356 , l53 bos , m. , retter , a. , cook , l. , & novak , r. 2001 , , 7665 , 2 brown , n. , & pearce , a. 2002 , , 7982 , 3 burlak , m. a. 2008 , astronomy letters , 34 , 249 burlak , m. a. , & henden , a. a. 2008 , astronomy letters , 34 , 241 burwitz , v. , starrfield , s. , krautter , j. , & ness , j .- u . 2002 , aip conf . proc .
637 , classical nova explosions , eds .
m. hernanz & j. jos ( aip , new york ) , 377 buscombe , w. , & de vaucouleurs , g. 1955 , the observatory , 75 , 170 capaccioli , m. , della valle , m. , rosino , l. , donofrio , m. 1989 , , 97 , 1622 capaccioli , m. , della valle , m. , donofrio , m. , rosino , l. 1990 , , 360 , 63 cassatella , a. , altamore , a. , & gonzlez - riestra , r. 2002 , , 384 , 1023 cassatella , a. , lamers , h. j. g. l. m. , rossi , c. , altamore , a. , gonzlez - riestra , r. 2004 , , 420 , 571 chochol , d. , grygar , j. , pribulla , komzik , r. , hric , l. , & elkin , v. 1997 , , 318 , 908 chochol , d. , hric , l. , urban , z. , komzik , r. , grygar , j. , & papousek , j. 1993 , , 277 , 103 cohen , j. g. 1985 , , 292 , 90 cohen , j. g. 1988 , the extragalactic distance scale , asp conference series vol . 4 ( astron . soc .
pacific : san francisco ) , 114 .
della valle , massimo ; livio , m. 1995 , , 452 , 704 della valle , m. , pasquini , l. , daou , d. , & williams , r. e. 1999 , , 7194 , 1 della valle , m. , pasquini , l. , daou , d. , & williams , r. e. 2002 , , 390 , 155 de vaucouleurs , g. 1978 , , 223 , 351 dobrotka , a. , friedjung , m. , retter , a. , hric , l. , novak , r. 2006 , , 448 , 1107 dobrotka , a. , retter , a. , & liu , a. 2008 , , 478 , 815 downes , r. a. , & duerbeck , h. w. 2000 , , 120 , 2007 drake , j. j. et al .
2007 , the astronomer s telegram , 1246 drake , j. j. et al .
2008 , the astronomer s telegram , 1721 drake , j. j. et al .
2003 , , 584 , 448 duerbeck , h. w. 1999 , inf .
stars , 4731 , 1 ennis , d. , beckwith , s. , gatley , i. , matthews , k. , becklin , e. e. , elias , j. , neugebauer , g. , & willner , s. p. 1977
, , 214 , 478 evans , et al .
2009 , , submitted ( arxiv : 0812.3662 ) eyres , s. p. s. , heywood , i. , obrien , t. j. , ivison , r. j. , muxlow , t. w. b. , & elkin , v. g. 2005 , , 358 , 1019 ferland , g. j. , lambert , d. l. , & woodman , j. h. 1986 , , 60 , 375 friedjung , m. 1987 , , 180 , 155 fujii , m. , et al . 1999 , , 7324 , 1 gallagher , j. s. , & ney , e. p. 1976
, , 204 , l35 gallagher , j. s. , kaler , j. b. , olson , e. c. , hartkopf , w. i. , & hunter , d. a. 1980 , , 92 , 46 gilmore , a. c. 1999 , iau circ . , 7179 , 1 goranskij , v. p. , metlova , n. v. , barsukova , e. a. , burenkov , a. n. , soloviev , v. ya .
2008 , the astronomer s telegram , 1631 hachisu , i. , & kato , m. 2001 , , 558 , 323 hachisu , i. , & kato , m. 2003 , , 590 , 445 hachisu , i. , & kato , m. 2004 , , 612 , l57 hachisu , i. , & kato , m. 2005 , , 631 , 1054 hachisu , i. , & kato , m. 2006 , , 167 , 59 hachisu , i. , & kato , m. 2007 , , 662 , 552 hachisu , i. , & kato , m. 2009 , , 694 , l103 hachisu , i. , kato , m. , & cassatella , a. 2008 , , 687 , 1236 hachisu , i. , et al .
2006 , , 651 , l141 hachisu , i. , kato , m. , & luna , g. j. m. 2007 , , 659 , l153 haseda , k. , west , d. , yamaoka , h. , & masi , g. 2002 , , 7975 , 1 heywood , i. , obrien , t. j. , eyres , s. p. s. , bode , m. f. , & davis , r. j. 2005 , , 362 , 469 hjellming , r. m. , van gorkom , j. h. , taylor , a. r. , seaquist , e. r. , padin , s. , davis , r. j. , & bode , m. f. 1986 , , 305 , l71 iijima , t. 2006 , , 451 , 563 iijima , t. , & esenoglu , h. h. 2003 , , 404 , 997 kaler , j. b. 1986 , , 98 , 243 kamath , u. s. , anupama , g. c. , ashok , n. m. , mayya , y. d. , sahu , d. k. 2005 , , 361 , 1165 kang , t. w. , retter , a. , liu , a. , & richards , m. 2006 , , 132 , 608 kato , m. , & hachisu , i. , 1994 , , 437 , 802 kato , m. , & hachisu , i. , 2007 , , 657 , 1004 kato , m. , hachisu , i. , & cassatella , a. 2009 , , 704 , 1676 kato , t. , et al .
2002 , , 7976 , 2 kawara , k. , maihara , t. , noguchi , k. , oda , n. , sato , s. , oishi , m. , & iijima , t. 1976 , , 28 , no . 1 , 1976 , p. 163 kiss , l. l. , & thomson , j. r. 2000 , , 355 , l9 krautter , j. , beuermann , k. , leitherer , c. , oliva , e. , moorwood , a. f. m. , deul , e. , wargau , w. , klare , g. , kohoutek , l. , van paradijs , j. , & wolf , b. 1984 , , 37 , 307 krautter , j. , gelman , h. , starrfield , s. , wichmann , r. , & pfeffermann , e. 1996 , , .456 , 788 lance , c. m. , mccall , m. l. , & uomoto , a. k. 1988 , , 66 , 151 liller , w. 2005 , , 8559 , 1 liller , w. , & jones , a. 2000 , inf .
stars , 5004 , 1 .
lockwood , g. w. , & millis , r. l. 1976 , , 88 , 235 lynch , d. k. , russell , r. w. , rudy , r. j. , & woodward , c. e. 2008 , , 8883 , 1 lynch , david k. , russell , r. w. , rudy , r. j. , & woodward , c. e. 2009 , american astronomical society , aas meeting 213 , 491.25 mallama , a. d. , & skillman , d. r. 1979 , , 91 , 99 mazuk , s. , lynch , d. k. , rudy , r. j. , russell , r. w. , pearson , r. l. , woodward , c. e. , & puetter , r. c. 2007 , iau circ .
, 8848 , 1 mazuk , s. , rudy , r. j. , lynch , d. k. , venturini , c. , puetter , r. c. , armstrong , t. 2000 , , 7490 , 3 morgan , g. e. ; ringwald , f. a. ; prigge , j. w. 2003 , , 344 , 521 mukai , k. , & ishida , m. 2001 , , 551 , 1024 munari , u. , dalla via , g. , valisa , p. , dallaporta , s. , & castellani , f. 2007 , cbet , 897 munari , u. , henden , a. , pojmanski , g. , dallaporta , s. , siviero , a. , & navasardyan , h. 2006a , , 369 , 1755 munari , u. , et al .
2008 , , 492 , 145 munari , u. , siviero , a. , navasardyan , h. , & dallaporta , s. 2006b , , 452 , 567 naik , s. , banerjee , d. p. k. , ashok , n. m. 2009 , , 394 , 1551 naito , h. , narusawa , s. , yamaoka , h. , & fujita , y. 2007 , , 8812 , 2 nakano , s. , kadota , k. , waagen , e. , swierczynski , s. , komorous , m. , king , r. , & bortle , j. 2007a , iau circ . , 8861 , 1 nakano , s. , tago , a. , nishiyama , k. , & sakamoto , t. 2007b , iau circ .
, 8821 , 1 nakano , s. , tago , a. , sakurai , y. , kushida , r. , & kadota , k. 2004 , iau circ .
, 8443 , 1 nelson , t. , et al .
2007 , the astronomer s telegram , 1202 ness , j .- u . , et al .
2003 , , 594 , l127 ness , j .- u .
2009 , , 137 , 4160 ness , j .- u . ;
schwarz , g. ; osborne , j. p. ; page , k. ; drake , j. j. , & krautter , j. 2008b , the astronomer s telegram , 1747 , 1 ness , j .- u . , schwarz , g. j. , retter , a. , starrfield , s. , schmitt , j. h. m. m. , gehrels , n. , burrows , d. , & osborne , j. p. 2007a
, , 663 , 505 ness , j .- u .
, starrfield , s. , jordan , c. , krautter , j. , & schmitt , j. h. m. m. 2005 , , 364 , 1015 ness , j .- u . ,
starrfield , s. , schwarz , g. , bode , m. f. , osborne , j. p. , & page , k. l. 2008a , the astronomer s telegram , 1370 ness , j .- u . ,
starrfield , s. , schwarz , g. , osborne , j. p. , & page , k. l. 2007b , the astronomer s telegram , 1327 , 1 ness , j .- u . , et al .
2008c , , 8911 , 2 ness , j .- u . , starrfield , s. , schwarz , g. , page , k. l. , & osborne , j. p. 2007c
, cbet , 1030 , 1 nielbock , m. , & schmidtobreick , l. 2003 , , 400 , l5 nomoto , k. 1982 , , 253 , 798 nomoto , k. , thielemann , f. , & yokoi , k. 1984 , , 286 , 644 novak , r. , retter , a. , cook , l. , korcakova , d. , & janik , j. 2000 , , 7448 , 4 orio , m. , covington , j. , gelman , h. 2001a , , 373 , 542 orio , m. , et al .
2003 , iau circ . , 8131 , 2 orio , m. , & tepedelenlioglu , e. 2003 , iau circ .
, 8435 , 2 orio , m. , parmar , a. n. , greiner , j. , gelman , h. , starrfield , s. , & trussoni , e. 2002 , , 333 , l11 orio , m. , et al . 2001b , , 326 , l13 pereira , a. , di cicco , d. , vitorino , c. , & green , d. w. e. 1999 , , 7323 , 1 pereira , a. j. s. , mcgaha , j. e. , young , j. , & rhoades , h. 2007 , , 8895 , 1 poggiani , r. 2008 , , 315 , 79 poggiani , r. 2009 , astron .
nachr . , 330 , 77 pojmanski , g. , szczygiel , d. , & pilecki , b. 2007 , iau circ . 8899 .
prialnik , d. , & kovetz , a. 1995 , , 445 , 789 read , a. m. et al .
2008 , , 482 , l1 read , a. m. et al .
2007 , the astronomer s telegram , no .
1301 retter , a. , cook , l. , novak , r. , saxton , j. m. , jensen , l. t. , korcakova , d. , & janik , j. 2000 , , 7537 , 2 rohrbach , j. g. , ness , j .- u . , & starrfield , s. 2009 , , 137 , 4627 rosino , l. 1987 , rs ophiuchi ( 1985 ) and the recurrent novae phenomenon , ed . m. f. bode ( utrecht , vnu science press ) , 1 rudy , r. j. , et al .
2006 , bull . of the american astr .
, 38 , 907 russell , r. w. , rudy , r. j. , lynch , d. k. , mazuk , s. , pearson , r. l. , woodward , c. e. , puetter , r. c. , & perry , r. b. 2007a , , 8846 , 1 russell , r. j. , rudy , r. j. , lynch , d. k. , & woodward , c. e. 2007b , iau circ . , 8888 , 1 sala , g. , hernanz , m. , & ferri , c. 2007 , the astronomer s telegram , 1184 sala , g. , hernanz , m. , ferri , c. , & greiner , j. 2008 , , 675 , l93 schmidt , th .
1957 , z. astrophys . , 41 , 181 selvelli , p. ; friedjung , m. 2003 , , 401 , 297 senziani , f. , skinner , g. k. , jean , p. , & hernanz , m. 2008 , , 485 , 223 shanley , l. , gelman , h. , gallagher , j. s. , orio , m. , & krautter , j. 1995 , , 438 , l95 shore , s. n. et al .
2003 , , 125 , 1507 siviero , a. , munari , u. , & jones , a. f. 2005 , inf .
stars , 5638 , 1 slavin , a. j. , obrien , t. j. , & dunlop , j. s. 1995 , , 276 , 353 snijders , m. a. j. 1987a , rs ophiuchi ( 1985 ) and the recurrent novae phenomenon , ed . m. f. bode ( utrecht , vnu science press ) , 51 snijders , m. a. j. 1987b , , 130 , 243 sostero , g. , yamaoka , h. , kiyota , s. , nakajima , k. , & maehara , h. 2004 , , 8445 , 3 starrfield , s. , ness , j .- u . , drake , j. j. , & orio , m. 2003 , , 8107 , 1 steiner , j. e. , & campos , r 1999 , , 7185 , 2 stickland , d. j. , penn , c. j. , seaton , m. j. , snijders , m. a. j. , & storey , p. j. 1981 , , 197 , 107 swierczynski , e. , ragan , e. , galan , c. , & mikolajewski , m. 2008 , the astronomer s telegram , 1723 tarasova , t. n. 2008 , odessa astronomical publications , 21 , 120 tempesti , p. 1979 , astron .
nachricheten , bd .
300 , 51 tomkin , j. , lambert , d. l. , & woodman , j. 1976 , , 48 , 319 torres , m. a. p. , jonker , p. g. , challis , p. , modjaz , m. , kirshner , r. , read , a. m. , kuulkers , e. , & saxton , r. d. 2007 , the astronomer s telegram , no .
1285 tomov , t. , et al .
2007 , inf .
stars , 5779 , 1 tsujimoto , m. , takei , d. , drake , j. j. , ness , j .- u .
, & kitamoto , s. 2008 , , 61 , s69 tylenda , r. 1978 , acta astronomica , 28 , 333 warner , b. , & woudt , p. a. 2009 , , 397 , 979 webbink , r. f. , rappaport , s. , & savonije , g. j. 1983 , , 270 , 678 wesson , r. et al .
2008 , , 688 , l21 whitelock , p. a. , carter , b. s. , feast , m. w. , glass , i. s. , laney , d. , menzies , j. w. , walsh , j. , & williams , p. m. 1984 , , 211 , 421 williams , p. , & gilmore , a. c. 1999 , iau circular no . 7176 , 1 woodward , c. e. , gehrz , r. d. , jones , t. j. , lawrence , g. f. , & skrutskie , m. f. 1997 , , 477 , 817 woodward , c. e. , wooden , d. h. , pina , r. k. , & fisher , r. s. 1999 , iau circ . , 7220 , 3 woudt , p. a. , warner , b. , & spark , m. 2005 , , 364 , 107 wu , c .- c .
, holm , a. v. , panek , r. j. , raymond , j. c. , hartmann , l. w. , & swank , j. h. 1989 , , 339 , 443 yamaoka , h. , nakamura , y. , itagaki , k. , nakano , s. , & nishimura , h. 2007 , , 8810 , 1 | on the basis of the recently developed universal decline law of classical novae , we propose prediction formulae for supersoft x - ray on and off times , i.e. , @xmath0 days and @xmath1 days for @xmath2 days . here
@xmath3 is the newly proposed `` intrinsic '' decay time during which the brightness drops by 3 magnitudes from optical maximum along our universal decline law fitted with observation .
we have determined the absolute magnitude of our free - free emission model light curves and derived maximum magnitude vs. rate of decline ( mmrd ) relations .
our theoretical mmrd relations are governed by two parameters , one is the white dwarf ( wd ) mass and the other is the initial envelope mass at a nova outburst ; this second parameter explains the scatter of mmrd points of individual novae .
our theoretical mmrd relations are also in good agreement with the well - known empirical formulae .
we also show another empirical relation of @xmath4 based on the absolute magnitude of our model light curves , i.e. , the absolute magnitude at 15 days after optical maximum is almost common among various novae .
we analyzed ten nova light curves , in which a supersoft x - ray phase was detected , and estimated their wd masses .
the models best reproducing simultaneously the optical and supersoft x - ray observations are onemg wds with @xmath5 ( v598 pup ) , @xmath6 ( v382 vel ) , @xmath7 ( v4743 sgr ) , @xmath8 ( v1281 sco ) , @xmath9 ( v597 pup ) , @xmath10 ( v1494 aql ) , @xmath11 ( v2467 cyg ) , @xmath12 ( v5116 sgr ) , @xmath13 ( v574 pup ) , and a co wd with @xmath14 ( v458 vul ) .
the newly proposed relationships are consistent with the emergence or decay epoch of the supersoft x - ray phase of these ten novae .
finally , we discuss the mechanism of shock - origin hard x - ray component in relation to the emergence of companion star from the wd envelope . |
the discovery of a new gauge boson has been reported by cms and atlas collaborations at the large hadron collider ( lhc ) @xcite . at present ,
its properties are found to be remarkably similar to standard model ( sm ) higgs @xcite .
to know more of its properties becomes a paramount task , since if we know it well , we may decide whether there is really new physics beyond sm and to determine further what s the new physics could be like .
the higgs decay channel , @xmath0 , which provides a dominant role for the discovery of higgs boson , shall be of great importance for revealing detailed properties of higgs boson at the lhc or future collider experiments .
the recent updates on higgs boson search show that the signal significance of @xmath0 channel are @xmath10 for cms @xcite and @xmath11 for atlas @xcite .
the decay width @xmath12 also enters into the cross section for higgs - boson production through photon - photon fusion , then , it could also be helpful for future high energy @xmath13 colliders as international linear collider ( ilc ) @xcite or the higgs factory .
therefore , it is important to study this decay channel as precise as possible . theoretically , within the sm model , the decay of higgs boson into photons is mediated though either @xmath14 boson or heavy fermions at least at the one - loop level , whose total decay rate can be generally written as @xmath15 where @xmath16 and @xmath17 with @xmath18 corresponding to top quark , bottom quark , charm quark and @xmath19 lepton , respectively .
the electroweak @xmath14 boson loop provides dominant contribution to the process , which is about @xmath20 times larger than that of the top - quark loop .
the fermionic contributions are proportional to an overall factor @xmath21 , which indicates that the fermionic contributions are dominated by top - quark loop and the contributions from the light fermions as @xmath22 or @xmath23 quarks can be safely neglected , especially for their higher loop corrections . because of its importance , in the literature
, many efforts have been made on studying the higgs decays into two photons , e.g. the one - loop estimation can be found in refs.@xcite , the two - loop estimation can be found in refs.@xcite , and the electroweak corrections at the two - loop level have also been analyzed in refs.@xcite .
most importantly , three - loop non - singlet qcd correction has been given in ref.@xcite , and recently , a complete three - loop correction including both singlet and non - singlet qcd corrections has been done by ref.@xcite .
all those improvements provide us a great chance for deriving more accurate estimation on higgs properties , as is the purposes of our present paper .
it is conventional to choose a typical momentum transfer of the process as the renormalization scale and take an arbitrary range to estimate the uncertainty in the qcd prediction .
as is well - known , under such conventional scale setting , there is renormalization scale and renormalization scheme ambiguities at any finite order .
it is often argued that by varying the renormalization scale , one can estimate contributions from higher - order terms .
however , this procedure only exposes the @xmath1-dependent nonconformal terms , not the entire perturbative series .
and the prediction of conventional scale setting is usually wrong in qed where there is never a scale uncertainty .
further more , the value of @xmath24 entering the qcd @xmath25-function is not determined by using conventional scale setting .
a recent review on this point can be found in ref.@xcite .
it has been suggested that the principle of maximum conformality ( pmc ) provides a possible solution for eliminating renormalization scale ambiguity @xcite , which also provides a principle to set the optimal renormalization scales up to all orders and to set the value of @xmath24 entering the qcd @xmath25-function at each perturbative order . in the present paper
, we shall adopt pmc to eliminate the renormalization scale ambiguity for @xmath0 . for the purpose
, one can first finish the renormalization procedure by using an arbitrary initial scale @xmath26 , and then set the effective or optimal pmc scales for the process by absorbing all non - conformal terms into the coupling constant via an order - by - order manner .
it is noted that the pmc satisfies all self - consistency conditions for setting the renormalization scale , and the qcd predictions under pmc are then independent of renormalization scheme .
because of these merits , the pmc method can be widely applied to high energy physics processes , some of its applications have already been shown in the literature @xcite . in principle , the scale dependence of the strong coupling constant is controlled by the renormalization group equation ( rge ) via the @xmath25 function , i.e. , @xmath27 where various terms in @xmath28 , @xmath29 , @xmath30 , correspond to one - loop and two - loop @xmath30 contributions respectively . if one can find a proper way to sum up all known - type of @xmath1-terms into the coupling constant , then one can determine the effective coupling for a specific process definitely at each perturbative order , and thus , the renormalization scale dependence can be greatly suppressed or even be eliminated . the pmc has been designed for such purpose . as an important point of deriving reliable pmc estimation ,
it is noted that only those non - conformal @xmath1-terms that rightly determine the running behavior of the coupling constant via rge should be absorbed into the coupling constant to form a conformal series . those unrelated @xmath1-terms that are from the quark mass renormalization and etc . should be kept as a separate during pmc scale setting . since all @xmath1-terms are entangled with each other
, one should find a proper way to deal with each types via a confidential way . in this paper , by suggesting an unambiguous way to carefully deal with the @xmath1-terms
, we shall make a detailed pmc analysis on higgs decay @xmath0 up to three - loop qcd corrections .
the remaining parts of this paper are organized as follows . in sec.[sec:2 ] , we present the calculation technology for the higgs decay @xmath0 up to three - loop qcd corrections under the pmc scale setting , in which we suggest a new way to set the pmc scales for high energy processes unambiguously . in sec.[sec:3 ] , we present the numerical results and discussions .
the sec.[sec:4 ] is reserved for a summary .
the pmc scales of the process can be determined by absorbing @xmath31-terms into the coupling constant . it is noted that the three - loop expressions presented in ref.@xcite are calculated under the @xmath2 scheme and are given for the top quark @xmath2 mass ( @xmath32 ) .
then , the @xmath1-terms for both the top - quark anomalous dimension and the qcd @xmath25-function are entangled with each other .
this makes it hard to set pmc scales in an unambiguous way . to solve the problem
, we suggest to transform expressions in the terms of @xmath32 into those of the top quark on - shell mass ( @xmath33 ) . after doing such transformation ,
all remaining @xmath1-terms are rightly pertained to the coupling constant and the pmc scales can be readily determined .
practically , this transformation can be achieved by using the relation between @xmath32 and @xmath33 @xcite .
more specifically , their relation up to @xmath34 can be written as @xmath35 .
\label{msmasspo}\end{aligned}\ ] ] after applying such transformation from @xmath36 to @xmath37 on the decay width for @xmath0 , we can rewrite the decay width into the following schematic form @xmath38 , \label{hrrconv}\ ] ] where @xmath39 stands for an arbitrary initial choice of renormalization scale , and under conventional scale setting , it is usually fixed to be typical momentum of the process , e.g. @xmath40 .
the coefficients , similar to ref.@xcite are defined as , @xmath41+\bigg(\hat a_t a_{t}^{(1)}(\tau_{t})\bigg)^{2},\\ a_{\rm ew}&= & 2\sqrt{a_{\rm lo}}a^{(1)}_{\rm ew } , \end{aligned}\ ] ] where @xmath42 .
the one - loop functions @xmath43 and @xmath44 , together with @xmath45 denoting the electroweak corrections to the w boson and top - quark induced processes , are free from strong coupling and should be kept as constants when applying pmc , whose explicit forms are @xcite @xmath46,\\ a_{f}^{(0)}(\tau_{f})&= & n_c\frac{\alpha \sqrt{\sqrt{2 } g_f}}{\pi \tau_{f } } q_f^2 \left[1+\left(1-{1\over \tau_{f}}\right ) f(\tau_{f})\right ] , \label{htolorrat}\end{aligned}\ ] ] where @xmath47 , @xmath48 is the fermi constant , @xmath49 denotes the quark electric charge and @xmath50 ^ 2 & \quad \mbox{for}\quad \tau>1 \end{array}\right . \;.\end{aligned}\ ] ] the @xmath44 with @xmath51 corresponding to charm quark , bottom quark and @xmath19 loop contributions at lo . for the case of @xmath19
, we should set @xmath52 .
the coefficients @xmath53 and @xmath54 are scale dependent , in which @xmath55 with @xmath56 are from one- , two- , and three - loop qcd contributions .
since the value of @xmath57 , we can expand @xmath55 in power series of @xmath58 .
the one - loop @xmath59 and two - loop @xmath60 can be expanded in the following forms up to order @xmath61 , @xmath62 the three - loop contribution @xmath63 includes two parts , one is the non - singlet part @xmath64 and the other is the singlet part , which can be further divided into @xmath65 for the terms with two top - quark loops and @xmath66 for the terms with one top - quark loop and one light - quark loop .
that is , @xmath67 where @xmath68 is the electromagnetic charge of the light quark @xmath69 . because the singlet parts @xmath65 and @xmath66 do not have @xmath31-terms ( or @xmath70-terms ) , so they are also kept as constant during the pmc scale setting .
their analytic expressions for @xmath71 can be derived from ref.@xcite . here , for brevity , we present @xmath65 and @xmath66 in numerical form by setting @xmath72 gev and @xmath73 gev , i.e. @xmath74 up to order @xmath61 , the non - singlet part @xmath64 in the terms of on - shell mass @xmath33 can be written as @xmath75 + \tau^{2}_{t } \bigg [ n_{f}\bigg(-{12504637\over34836480 } - { 4 \pi^2\over189}\nonumber\\ & & - { 4432\over42525}\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2 } } + { 61397 \zeta_{3}\over221184}\bigg ) + \bigg(-{51082579973\over2612736000 } + { 4 \pi^2\over27 } + { 8 \pi^{2}\over189 } \ln2 \nonumber\\ & & + { 24376 \over14175}\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2 } } + { 66912377 \zeta_{3}\over 3317760}\bigg ) \bigg]+ \tau^{3}_{t } \bigg [ n_{f } \bigg(-0.3364 - { 26 \pi^2\over 1575 } \nonumber\\ & & -0.0703\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2}}+0.2852 \zeta_{3}\bigg ) + \bigg(-827.988 + { 26 \pi^2\over 225}+ { 52 \pi^2\over 1575}\ln2 \nonumber\\ & & + 1.1595 \ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2 } } + 691.894\zeta_{3}\bigg ) \bigg]+ \tau^{4}_{t } \bigg [ n_{f } \bigg(-0.3133 - { 2048 \pi^2\over 155925 } -0.0496\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2 } } \nonumber\\ & & + 0.2897\zeta_{3}\bigg ) + \bigg(-2408.83 + { 2048 \pi^2\over22275 } + { 4096 \pi^2 \over 155925 }
\ln2 + 0.8191\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2}}\nonumber\\ & & + 2006.25 \zeta_{3 } \bigg ) \bigg ] + \tau^{5}_{t } \bigg [ n_{f } \bigg(-0.2949 - { 6080 \pi^2\over567567}-0.0367 \ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2 } } + 0.2930\zeta_{3}\bigg ) \nonumber\\ & & + \bigg(-17838.5 + { 6080 \pi^2\over 81081 } + { 12160 \pi^2 \over 567567 }
\ln2 + 0.6049\ln{(\mu^{\rm init}_{r})^{2}\over m_{t}^{2}}+14841.7 \zeta_{3}\bigg ) \bigg],\end{aligned}\ ] ] where @xmath76 and @xmath77 stand for non - conformal and conformal terms of @xmath64 , respectively . with the help of the above expressions , we obtain the conformal and non - conformal terms for the three - loop coefficient @xmath54 , i.e. @xmath78 where
the non - conformal @xmath79 and the conformal @xmath80 can be written as @xmath81 + \bigg(\hat a_t a_{t}^{(1)}(\tau_{t})\bigg)^{2}.\end{aligned}\ ] ] after applying the standard pmc scale setting procedures to the decay width ( [ hrrconv ] ) , especially the @xmath24 terms are absorbed into the coupling constant with a combined form of @xmath82 , the decay width can be simplified as , @xmath83 , \label{htorrpmc}\end{aligned}\ ] ] where the pmc scale @xmath84 . \label{htorrpmcscale}\end{aligned}\ ] ] it is noted that at the present three - loop level , we have only @xmath28-terms , so we have one pmc scale .
it is noted that only the initial scale dependent logarithmic terms should be absorbed into the coupling constant simultaneously with the non - conformal terms via rge , and the remaining conformal coefficients are still at the initial scale @xmath39 .
more over , as will be shown later , the pmc scale @xmath85 only formally depends on the choice of the initial renormalization scale , its own value and hence the decay width are almost independent of the initial choice of renormalization scale , and then the renormalization scale dependence can be greatly suppressed or even eliminated at the three - loop level .
to do numerical calculation , we take higgs mass @xmath72 gev as its central value , the w boson mass @xmath86 gev and @xmath19 lepton mass @xmath87 gev , the fermi constant @xmath88 , and @xmath89 . we adopt two - loop @xmath90 running with the fixed point @xmath91 @xcite to set its corresponding @xmath92 , i.e. , we obtain @xmath93 gev , @xmath94 gev , @xmath95 gev and @xmath96 gev . by adopting the on - shell quark masses @xmath73 gev , @xmath97 gev and @xmath98 gev @xcite , we obtain the @xmath2-running masses @xmath99 gev , @xmath100 gev and @xmath101 gev . .decay
width for the decay @xmath102 up to three - loop level , in which @xmath103 . for the conventional scale
setting , the renormalization scale is fixed to be @xmath104 .
@xmath105 , @xmath106 , @xmath107 and @xmath108 denote the decay widths at lo , nlo , nnlo levels and qed correction , respectively .
@xmath109 stands for the total decay width up to nnlo level . [
cols="^,^,^,^,^,^ " , ] when applying the pmc scale setting , after resumming all @xmath28-terms into the coupling constant , the results will be much more steady over the scale changes . as a comparison ,
the scale dependence for the nlo decay width @xmath106 are presented in table [ scaleun ] , where three scales @xmath110 , @xmath7 , @xmath111 are adopted .
the decay width of the nlo qcd corrections under the conventional scale setting shows a strong dependence on @xmath104 , i.e. the scale errors for the nlo correction terms are @xmath112 $ ] for @xmath113 $ ] .
furthermore , under the conventional scale setting , the nnlo decay width also shows a much larger scale uncertainty , i.e. @xmath114 kev for @xmath115 $ ] .
in contrast , under the pmc scale setting , the decay width of nlo qcd corrections is almost unchanged .
this is because that the pmc scale itself is fixed and highly independent of @xmath104 . for @xmath102 under conventional scale setting and pmc scale setting .
the solid and dashed lines stand for the total decay widths up to nlo level and nnlo level , respectively.,title="fig:",scaledwidth=48.0% ] for @xmath102 under conventional scale setting and pmc scale setting .
the solid and dashed lines stand for the total decay widths up to nlo level and nnlo level , respectively.,title="fig:",scaledwidth=48.0% ] more explicitly , by using eq.([htorrpmcscale ] ) , we find that the pmc scale is fixed to be @xmath116 gev @xmath117 for any choice of @xmath104 .
this shows that the pmc scale is larger than the typical momentum flow of the process , i.e. the higgs mass @xmath7 , to a certain degree .
this larger scale is consistent with the above observations that after pmc scale setting the nlo correction to the decay width becomes smaller than that of the conventional scale setting .
this point can be shown explicitly in fig.([plot : sqrst ] ) , in which the solid and dashed lines stand for the total decay widths up to nlo level ( @xmath118 ) and nnlo level ( @xmath119 ) , respectively .
the pmc estimation as shown by the right diagram of fig.([plot : sqrst ] ) shows that the decay widths at each perturbative order are almost flat versus @xmath104 . in principle , at different pqcd orders , we need to set different pmc scales @xcite . at present , the nnnlo calculation is not available , then we have no @xmath31-terms to determine the pmc scale for nnlo terms , so we set the pmc scale for nnlo terms to be equal to the pmc scale for nlo terms , i.e. @xmath120 , which is calculated by eq.([htorrpmcscale ] ) .
this treatment will cause residual scale dependence after pmc scale setting @xcite . to provide a relatively reliable estimation on such residual scale dependence , following the idea of pmc , we first rewrite the nnlo coupling constant in eq.([htorrpmc ] ) as follow @xmath121
we observe that such newly log - term at one - order higher , corresponding to part of the nnnlo level , can compensate the scale changes at the nnlo level and result in an improved scale error in comparison to the conventional way @xcite .
thus , after pmc scale setting , we can eliminate the renormalization scale ambiguity and obtain a more accurate predications in comparison to the previous estimations for the higgs decay @xmath102 @xcite . as a comparison , we present the decay widths at each perturbative order under the conventional scale setting with pole top - quark mass and under the pmc scale - setting in the following .
that is , by using eqs.([hrrconv],[htorrpmc ] ) , we have @xmath122\ ; { \rm kev } , \label{hrrmspo}\\ \gamma(h\rightarrow \gamma\gamma)|_{\rm pmc}&= & \left[9.502 + 1.487\times10^{-1}+1.415\times10^{-2}\right]\ ; { \rm kev } , \label{hrrmspmc}\end{aligned}\ ] ] where the errors in eq.([hrrmspo ] ) is caused by varying initial scale @xmath39 from @xmath123 to @xmath111 . under the conventional scale
setting , the decay width of the nnlo terms increases with the increment of the scale , while the decay width of the nlo terms decreases with the increment of the scale , then as a combination , the total decay width up to the nnlo level varies slightly with the change of scale as shown in fig.([plot : sqrst ] ) .
after pmc scale setting , the residual scale dependence for the nnlo terms is less than @xmath124 , so our pmc estimation ( [ hrrmspmc ] ) shows no scale dependence even at the present three - loop order . as a final remark . for a pqcd estimation ,
it is helpful to predict what s the unknown " qcd corrections could be and how it will be changed with the improved qcd corrections . the conventional estimation done by varying
the scale over a certain range is not proper , since it can only estimate the non - conformal contribution but not the conformal one .
more over , after the pmc scale setting , the determined pmc scales are optimal and can not be varied simply , otherwise , it will explicitly break the renormalization group invariance and lead to unreliable estimation . to achieve an estimation of how the unknown " qcd corrections could be from the known " qcd corrections , we rewrite the decay width as @xmath125 . \label{hrrconvnew}\ ] ] up to @xmath126-loop qcd correction , @xmath127 , where @xmath128 and the known coefficients @xmath129 for the convention and pmc scale settings can be read from eqs.([hrrconv],[htorrpmc ] ) .
for example , under the conventional scale setting , @xmath130 and @xmath131 .
we suggest the following more conservative method for the scale error analysis or to estimation how the unknown " higher - order qcd corrections could be ; i.e. to take the scale uncertainty as the last known perturbative order .
more explicitly , the pqcd uncertainty at the @xmath132-order is @xmath133 , where both @xmath134 and @xmath135 are calculated by varying the initial scale to be within the region of @xmath136 $ ] and the symbol `` max '' stands for the maximum value of @xmath137 within this region .
this treatment is natural for pmc , since after pmc scale setting , the pqcd convergence is ensured and the only uncertainty is from the last term due to the unfixed pmc scale at this particular order .
we obtain @xmath138 and @xmath139 for the conventional scale setting ; @xmath140 and @xmath141 for pmc .
the error bars provide a consistent estimate of the unknown " qcd corrections under various scale settings ; i.e. , the exact value for the unknown " @xmath142 are well within the error bars predicted from @xmath143 .
a detailed discussion on how to estimate the unknown " qcd corrections for various scale settings from @xmath144 and @xmath145 up to four - loop level shall be presented elsewhere @xcite .
we have applied the pmc scale setting to study the decay width of @xmath102 up to three - loop level .
after the pmc scale setting , we obtain a renormalization scale and renormalization scheme independent estimation .
then , a more accurate estimation has been achieved .
more explicitly , * the pmc renormalization scale is formed by absorbing the @xmath31-terms that govern the running behavior of the coupling constant into the coupling constant . at present ,
the decay width has been calculated up to three - loop level by using the @xmath2 renormalization scheme .
because the @xmath146-terms for both the top - quark anomalous dimension and the qcd @xmath25-function are entangled with each other , it is hard to apply pmc unambiguously without knowing each part well .
thus , to separate the @xmath31-terms in an unambiguous and convenient way , we transform the expressions with the @xmath2-mass @xmath36 to those with the pole - mass @xmath33 before applying the pmc scale setting .
it is noted that , our present suggestion of dealing with the @xmath1-terms can also be extended to deal with other source terms involving @xmath1-series that are also unrelated to determine the running behavior of the coupling constant in order to derive a reliable pmc estimation . *
as shown by the left diagram of fig.([plot : sqrst ] ) , under the conventional scale setting , the nlo decay width @xmath106 shows a strong dependence on @xmath104 , one can obtain a convergent and less scale dependent estimations up to three - loop level .
especially , there is large scale cancelation between different perturbative terms , i.e. the decay width of the nnlo terms increases while the decay width of the nlo terms decreases with the increment of the scale .
in contrast , as shown by the right diagram of fig.([plot : sqrst ] ) , after the pmc scale setting , the decay widths at the each perturbative order are almost flat versus the initial renormalization scale @xmath104 .
it shows that the optimal renormalization scale for @xmath147 is @xmath4 other than the usual adopted @xmath148 .
it is noted that because the nlo level pmc scale can be definitely determined by the @xmath149-terms at the nnlo level , the total decay width up to nlo level is almost flat as shown by table [ scaleun ] ; while there is no @xmath1-terms to set the pmc scale for nnlo terms , so there is residual scale dependence for nnlo terms , which is quite small via an improved conventional scale setting method as suggested in ref.@xcite .
+ versus the higgs mass @xmath7 after the pmc scale setting .
, scaledwidth=48.0% ] * since the scale dependence has been eliminated , after the pmc scale setting , a more accurate scale - independent pqcd predication for @xmath102 can be obtained . by taking the region for the higgs mass as the same as the ones determined by atlas and cms collaborations ,
the decay width under the pmc scale setting are as follows , @xmath150 where the subscript @xmath151 means the error is caused by varying higgs mass @xmath152 gev determined by the atlas collaboration and the subscript @xmath153 means the error is caused by varying higgs mass @xmath154 gev determined by the cms collaboration , respectively .
more explicitly , we show the decay width versus the higgs mass @xmath7 in fig.([plot : sqpms : c ] ) .
in addition to previous examples done in the literature , the pmc scale setting works well for the decay channel @xmath102 .
inversely , if taking the decay width as an input , one can obtain a more accurate estimation on the higgs mass .
thus , it is helpful to reveal the properties of higgs boson with high precision , and it can also increase our understanding of possible new physics beyond the sm . it is noted that a four - loop estimation of @xmath147 has been presented very recently @xcite .
it , being given with top - quark @xmath2-mass , can not be directly dealt with by pmc .
however , we can estimate that the four - loop terms shall not affect our main conclusions since our present three - loop estimation already shows good pqcd convergence and it is reasonable to estimate that the four - loop terms shall provide small contribution after applying pmc scale setting . *
acknowledgments * : we thank stanley j. brodsky and matin mojaza for helpful discussions .
this work was supported in part by natural science foundation of china under grant no.11275280 , by the program for new century excellent talents in university under grant no.ncet-10-0882 , and by the fundamental research funds for the central universities under grant no.cqdxwl-2012-z002 . | the principle of maximum conformality ( pmc ) provides a convenient way for setting the optimal renormalization scales for high - energy processes , which can eliminate the conventional renormalization scale error via an order - by - order manner . at present , we make a detailed pmc analysis on the higgs decay @xmath0 up to three - loop qcd corrections . as an important point of deriving reliable pmc estimation , it is noted that only those @xmath1-terms that rightly determine the running behavior of coupling constant via the renormalization group equation should be absorbed into the coupling constant , and those @xmath1-terms that pertain to the quark mass renormalization and etc .
should be kept as a separate . to avoid confusion of separating and absorbing different types of @xmath1-terms into the coupling constant , we first transform the decay width in terms of top quark @xmath2 mass into that of on - shell mass and then apply the pmc scale setting . after applying pmc scale setting , the final estimation is conformal and is scheme - independent and scale - independent .
up to three - loop qcd corrections , we obtain a pmc scale @xmath3 gev @xmath4 , which is optimal and highly independent of any choice of initial scale .
thus , we obtain a more accurate scale - independent prediction by taking the higgs mass as the same as that of atlas and cms measurements , i.e. , @xmath5 kev and @xmath6 kev , where the error is caused by the measured higgs mass , i.e. the higgs mass @xmath7 is taken as @xmath8 gev for atlas and @xmath9 gev for cms , respectively .
+ pacs number(s ) : 12.38.bx , 14.80.bn , 11.10.gh |
gamma - ray burst ( grb ) pulse characteristics remain poorly understood even though pulses are the most common structures in grb light curves : there is little agreement concerning the physical mechanisms responsible for producing them .
pulse characteristics are difficult to extract due to a mixture of strong spectral - temporal evolution , instrumental and sampling biases , confusion caused by overlapping pulses , and low signal - to - noise measurements .
understanding pulse behaviors is important , as it is difficult to adequately model grb progenitors and outflow without first understanding the mechanism by which pulses liberate energy , and it is difficult to understand pulse mechanisms without first identifying pulse behaviors .
data driven analysis approaches are providing new insights into grb pulse characteristics .
these approaches have both advantages and disadvantages over standard theoretical modeling techniques : they can identify behaviors ( typically clustering and/or correlations ) that theoretical models are not trying to explain and thus may not recognize , but by asking questions about pulse behaviors they do not necessarily couch these questions in the form of specific theoretical models .
data driven approaches have allowed the identification of important pulse characteristics even as theoretical models have been unable to explain these behaviors . as an example , consider empirical light curve models , which can be used to provide an accurate representation of grb pulse behaviors in the absence of identified pulse physics .
several empirical models are able to adequately explain the general shapes of grb pulse light curves ( e.g. see the introduction in @xcite ) . among these ,
the simple model of @xcite is useful because it has only four free parameters which allow reasonable fits over the wide range of observed pulse characteristics .
empirical approaches demonstrate that grb pulse light curves can be generally represented by a _ hard - to - soft _ spectral evolution .
pulses are generally hardest at the first instant they appear , and they continue to soften during the pulse rise , past the pulse peak , and through the pulse decay @xcite .
although hard - to - soft evolution is the most common pulse evolution ( e.g. @xcite ) , _ intensity tracking _ pulse evolution has been identified in a smaller number of pulses ( e.g. @xcite ) .
it is difficult to estimate what fraction of pulses exhibit intensity tracking behaviors , since there are many cases where overlapping hard - to - soft pulses appear to have been misclassified as intensity tracking @xcite .
hard - to - soft spectral evolution is so central to grb pulse structure that it can be used as a defining characteristic in rigorous statistical modeling techniques @xcite .
the harder a pulse is , the more pronounced the evolution typically is .
a small percentage of pulses appear to harden during the pulse rise , but these _ intensity tracking _
pulses are typically among the very softest , suggesting that instrumental biases might be in part responsible for the initial soft emission . several of the well - known correlated spectral - temporal grb pulse properties are simply a result of this hard - to - soft spectral decay : rapid hardness evolution produces short duration pulses with short spectral lags while slow hardness evolution produces long duration ones with long lags .
the relationship between pulse asymmetry and hard - to - soft evolution is not as clear as the one between duration , lag , and evolution .
asymmetry is anti - correlated with hardness and correlated with duration across a large batse grb pulse sample such that asymmetric pulses are generally longer and softer than symmetric pulses @xcite .
this implies that asymmetric pulses in general do not undergo a significant hardness evolution , since the softening is mild and occurs over a long time interval .
however , some of the faintest , softest , longest grb pulses are symmetric and undergo very little hard - to - soft evolution , while some asymmetric pulses are quite hard . to further confuse matters , the bright , short pulses in classical short grbs are also symmetric yet very hard ; undergoing a rapid hard - to - soft evolution . without an obvious connection between hard - to - soft evolution and pulse asymmetry , the possibility exists that asymmetry and hard - to - soft evolution are independent or semi - independent pulse characteristics .
grb pulses ( at least those belonging to the long and intermediate grb classes ) exhibit additional intensity fluctuations that overlay the hard - to - soft monotonic pulse component @xcite ; these fluctuations can be seen in the residuals obtained by subtracting the @xcite modeled monotonic pulse shape from the light curve .
these fluctuations , which have the appearance of three peaks separated by two valleys , are more easily observed in high signal - to - noise pulses ; this makes bright pulses look like they have variable , multi - peaked light curves , in contrast to the smooth monotonic light curves of fainter pulses .
when combined with spectral softening , bright pulses appear to be composed of at least three peaks , with the first one ( during or immediately preceding the pulse rise ) generally being the hardest , the second ( and brightest ) one being of intermediate hardness , and the last one ( during the long decay ) being soft .
hard - to - soft evolution and phased non - monotonic intensity fluctuations can be used to help determine if a grb emission episode is composed of a single pulse or of several merged pulses . a single pulse , at least in the energy ranges of batse and fermi , has pulse residual parameters that correlate with the pulse asymmetry .
the residual light curve waves can be fitted by a simple four - parameter empirical model @xcite this empirical model has several interesting features : 1 ) the basis of the model is a truncated bessel function of the first kind , @xmath0 that begins at the true pulse peak , 2 ) the wave can be modeled by a mirror image of itself in time , with the time - reflected part occurring during the pulse rise and the time - forward part occurring during the pulse decay , 3 ) the pulse decay portion is a time - stretched version of the mirrored part on the pulse rise , and 4 ) most if not all of the pulse residual fit parameters correlate with the pulse asymmetry . at the present time , no physical mechanism is known that can provide these observed model characteristics .
it is the reproducibility of the light curve residual fluctuations , combined with the overall hard - to - soft spectral evolution , which demonstrates that the fluctuations all combine to form a single evolving pulse with a non - monotonic shape .
all pulse light curves seem to undergo multiple intensity increases and decreases
. however , the definition of a pulse being composed of several monotonically rising and falling regions may be a semantic one : a single pulse might indeed be composed of three separate , linked , smaller pulses .
if we do choose to believe that each fluctuation identifies a separate pulse , then the coupling between these fluctuations and the overall hard - to - soft evolution indicates that multiple pulses always combine to form similar , temporally - linked , complex shapes that need to be described by a single smooth hardness evolution .
the grbs used to measure pulse residual fluctuations have thus far all been those observed by batse and fermi s gbm experiment .
since these instruments have spectral responses favoring detection of higher energy photons ( 25 kev to 2 mev ) over lower ones , it is reasonable to wonder whether these residual features are present at lower energies observed by swift , or if they are primarily a high energy phenomenon . to this end
, we examine a sample of single episodic bursts observed by swift .
the swift grbs analyzed here were observed between august 1 , 2005 and september 1 , 2009 , and were selected by eye from a sample having measured redshifts as being those having single emission episodes .
the initial sample contains 30 grbs .
similar to batse , single - pulsed bursts appear to comprise roughly one - third of all swift grbs ( one - quarter if the sample is limited to to long bursts ) .
light curves are composed of 64 ms , mask - weighted , background - subtracted , 4-channel data available from swift .
the swift four - channel data are similar to those used for batse and fermi gbm , but optimized to the swift energy range ( channel 1 @xmath1 kev , channel 2 @xmath2 kev , channel 3 @xmath3 kev , and channel 4@xmath4 kev ) .
the technique used for extracting pulses from grb light curves has been described elsewhere ( e.g. @xcite and references contained therein ) .
intervals in the light curve where pulses may reside are first identified using the bayesian blocks approach of @xcite , followed by a pulse - fitting technique that fits the empirical model of @xcite to each interval using a nonlinear least squares routine @xcite . for batse and fermi gbm data ,
poisson statistics are assumed . for swift data ,
intensity uncertainties are gaussian , and this requires a modification of the bayesian blocks code ( norris 2014 , private communication ) .
the @xcite pulse model is defined as @xmath5},\ ] ] where @xmath6 is time since trigger , and the four norris pulse fit parameters are the pulse amplitude @xmath7 , the pulse start time @xmath8 , the pulse rise parameter @xmath9 , and the pulse decay parameter @xmath10 .
the normalization constant @xmath11 is defined in terms of these parameters as @xmath12}$ ] .
similarly , the pulse peak time occurs at time @xmath13 . in obtaining the pulse fit , the start and end time values of each bayesian block
are used to obtain initial estimates of @xmath8 , @xmath9 , and @xmath10 , while the maximum amplitude occurring during the block is used * to * estimate @xmath7 .
observable pulse parameters are defined from the four free pulse fit parameters .
we have previously defined the pulse duration @xmath14 as the time interval between instances at which the intensity is @xmath15 ( @xmath16 ) of its maximum value @xcite , @xmath17^{1/2},\ ] ] where @xmath18 .
the pulse _ asymmetry _ @xmath19 is also defined from the intensities at @xmath15 of the maximum value , and it has a value of @xmath20^{-1/2};\ ] ] it ranges from a value of @xmath21 for a symmetric pulse to @xmath22 for an asymmetric pulse with a rapid rise and slow decay .
the @xcite model provides a good general fit to batse and fermi gbm pulses , but close examination have identified small yet distinct deviations in the residuals that are phased to the pulse duration @xcite .
these variations have been found to be contained within the fiducial time interval @xmath23 defined by : @xmath24 with the fiducial end time @xmath25 given by @xmath26 and the fiducial start time @xmath27 given by @xmath28.\ ] ] in the fiducial time interval , the residual variations can be fitted with the empirical four parameter function described by @xcite : @xmath29 } ) & \textrm{if $ t < t_0 - 0.005 $ } \\ a & \textrm{if $ t_0 - 0.005 \le
t \le t_0 + 0.005 $ } \\ a j_0(\sqrt{s \omega [ t - t_0 - 0.005 ] } ) & \textrm{if $ t >
t_0 + 0.005$. } \end{array } \right.\ ] ] here , @xmath0 is an integer bessel function of the first kind , @xmath30 is the central time of the residuals peak , @xmath31 is the amplitude of the residual peak , @xmath32 is the variability factor the bessel function s angular frequency which defines the timescales of residual wave ( a large @xmath32 corresponds to a rapid rise and fall ) , and @xmath33 is a scaling factor that relates the fraction of time which the function before @xmath30 has been compressed relative to its time - inverted form after @xmath30 ( or a stretching of the function after the peak time relative to that before it ) . using this model , the time during which the pulse intensity is a maximum needs to be a plateau instead of a peak , with a duration of @xmath34 . since there is no evidence that the bessel function continues beyond the third zero ( following the second half - wave ) , the function is truncated at the third zeros @xmath35 . in case
it is not intuitively obvious what roles each of these four free parameters plays in the residual fit , we have included examples in which we vary each parameter individually within the fiducial time interval .
the results are shown in figure 1a ( varying @xmath31 ) , figure 1b ( varying @xmath30 ) , figure 2a ( varying @xmath32 ) , and figure 2b ( varying @xmath33 ) . for these residual models
we have assumed that the underlying pulse has an asymmetry of @xmath36 , and that the fiducial timescale is defined to encapsulate a pulse of this asymmetry .
the fiducial values can be converted back to measured values using the following relationships : @xmath37 @xmath38 @xmath39 and @xmath40 where @xmath27 and @xmath25 are the real time values corresponding to the start and end of the fiducial duration .
a convenient way to describe the pulse residual amplitudes is to normalize them to the pulse fit amplitudes , because the result is independent of the instrument s signal - to - noise , allowing amplitude comparisons between different instruments .
this produces a relative amplitude @xmath41 defined by @xmath42
we apply these pulse - fitting techniques to our sample of 30 swift grbs having single emission episodes , with our initial hypothesis being that each emission episode represents a single pulse .
we feel justified in making this assumption because the @xcite results suggest that the residual pattern underlies all isolated pulse structures .
this hypothesis is testable ; a pulse exhibiting the residual wave structure should have residual wave properties that correlate with the pulse asymmetry in the manner found previously for batse pulses .
thus , we fit the swift 4-channel grb data using the @xcite model and the techniques described by @xcite , but we increase the ncp prior variable ( a prior value for the number of change points occurring in the light curve ) in the bayesian blocks code @xcite and/or the level of dual timescale peak flux significance @xcite so that only a single fitted pulse is returned for each grb in the sample . even though the swift grb sample has been selected from what appear to be bright single - pulsed bursts observed during a predetermined time period , many of the light curves are relatively noisy with large uncertainties .
this is not surprising , as grb emission typically decreases at lower energies where background noise increases . as a result
, the four - channel pulse fits are generally not as good as those obtained for batse and fermi gbm data , and two bursts ( grb 080129 and grb 090313 ) can not be fitted using the @xcite pulse model .
the results of the @xcite fits to the other 28 pulses in the swift sample are summarized in table 1 .
we attempt to fit the remaining 28 pulses with both the @xcite model and the @xcite residual fit model .
we use a _ @xmath43 difference test _
@xcite to determine if the fit has been improved with the addition of the @xcite residual fit . in this test
, the difference in the @xmath43 fit parameters between the two models @xmath44 is given as @xmath45 where the difference in degrees of freedom @xmath46 is @xmath47 here ` p ' refers to the pulse fit and ` r ' refers to the residual fit .
a resulting probability of @xmath48 is obtained for a @xmath44 value having @xmath46 degrees of freedom .
pulse plus residual fits having @xmath49 do not show significant improvement with the additional model parameters and are excluded , while pulse plus residual fits having @xmath50 are significant enough to be retained . as shown in table 2 ,
only eleven pulses in our swift sample satisfy the @xmath50 criterion : grb 051111 , grb 060206 , grb 060708 , grb 060912a , grb 061004 , grb 061021 , grb 070318 ,
grb 070117 , grb 080413b , grb 081222 , and grb 091018 .
the residual fits and combined fits of these pulses are summarized in table 3 and are shown in figures 3 through 13 ( panels on the left side indicate the fit to the residuals while panels on the right side indicate the overall fit ) . in order to compare the results of
the swift fits with those previously obtained for batse and fermi pulses , we also apply the @xmath43 difference test to the pulses listed in @xcite .
six batse pulses ( batse 0673 , batse 0711 , batse 0727 , batse 1301 , batse 7567 , and batse 7775 ) and one fermi pulse ( grb 120326a ) are not found to exhibit @xmath44 fit improvements and have been subsequently excluded from the analysis .
it seems possible that _ all _ grb pulses might be characterized by a pulse model that includes the same residual wave pattern found in the batse data , but that this residual pattern becomes harder to measure for faint pulses . to test this hypothesis on our isolated pulse sample
, we check to see if the residual pattern is easier to identify when pulses are bright .
we use the signal - to - noise ratio @xmath51 defined as the peak 1 s pulse intensity divided by the peak 1 s pulse intensity uncertainty ; these intensities are available for swift in table 1 and for batse and gbm from published online grb catalogs .
the @xmath51 values depend on both instrumental sensitivity and instrumental energy response . in figure 14
we compare our measure of the fit improvement ( @xmath52 $ ] ) to the @xmath51 for each pulse ( a large @xmath53 value represents an improved fit ) .
a spearman rank order test applied to batse pulses ( crosses ) yields a p - value of @xmath54 that @xmath55 and @xmath51 are uncorrelated while a similar test applied to swift pulses ( filled circles ) yields a p - value of @xmath56 .
the trend also appears to hold for the three fermi pulses ( open diamonds ) .
these results demonstrate that the fits improve significantly as the signal - to - noise increases , which is consistent with our hypothesis .
the pulses not substantially improved by the residual fits also support the hypothesis that the residual wave is present in all isolated pulses : these unimproved pulses are typically among the faintest pulses in the sample .
the mean @xmath51 of the pulses whose fits have been improved ( @xmath50 ) is @xmath57 while the the mean @xmath51 of the unimproved pulses ( @xmath49 ) is @xmath58 a student s t - test indicates only a @xmath59 likelihood that these distributions have the same mean .
the characteristics of the wave patterns extracted from the improved swift grb pulse residuals are found to be similar to those of batse grb pulse residuals in that they exhibit similar correlations with pulse asymmetry @xmath19 .
figure 15 demonstrates the relationship between the variability parameter @xmath32 with @xmath19 , figure 16 demonstrates the relationship between the expansion parameter @xmath33 and @xmath19 , and figure 17 demonstrates the relationship between the fiducial residual peak time @xmath30 and @xmath19 .
these correlations show a clear linkage between the shape of the episode s bulk emission and the more rapidly - varying component contained within the episode .
since the peak of the residuals @xmath30 occurs close in time to the fitted @xcite pulse peak @xmath60 , both @xmath30 and @xmath60 correlate with @xmath19 .
since the residual peaks are closely associated with pulse rise and decay , the stretching parameter @xmath33 correlates with @xmath19 . and
, since an asymmetric pulse has a short rise time relative to a longer decay time , it has a large variability and therefore a large @xmath32 to go with a large @xmath19 .
there is no observed correlation between @xmath41 and @xmath19 , nor is there any clear correlation between @xmath41 and @xmath51 .
figure 18 shows that the mean value of @xmath41 is @xmath61 , which indicates that the residual peak intensity @xmath31 is generally @xmath62 of the fitted peak intensity @xmath7 ; this is true for all pulses in which the residuals can be fitted , regardless of the instrument making the measurements .
pulse residuals describing the pulse shape are just as identifiable in symmetric pulses as they are in asymmetric pulses .
this does not mean that symmetric pulses are as easy to observe as asymmetric pulses : the fitted pulse amplitude @xmath7 has been found to correlate with @xmath19 for batse pulses @xcite , indicating that asymmetric pulses are generally brighter than symmetric pulses .
the residual pattern is not only easier to detect in bright pulses ; it is also easier to detect at higher energies , in pulses with intrinsically larger @xmath63 values , and in asymmetric pulses ( @xmath64 .
figure 19 contrasts the characteristics of pulses that have not been improved by the residual fits ( @xmath65 ; in blue ) with the characteristics of pulses that have been improved by the residual fits ( @xmath66 ; in black ) .
improved swift fits generally require higher @xmath51 ratios than improved batse fits . since batse and swift pulses have similar minimum @xmath51 ratios indicative of comparable instrumental sensitivities , this difference between residual intensities most likely results from the different spectral responses of the instruments , and suggests that the residual structures are more apparent at high energies than at low energies .
the left panel of figure 19 demonstrates an anti - correlation between the normalized residual amplitude @xmath63 and the signal - to - noise @xmath51 : faint pulses with small @xmath51 ratios tend to have larger values of @xmath41 .
unimproved pulses near the threshold tend to balance this anti - correlation somewhat ( this is especially true for batse pulses ) by having smaller @xmath41-values .
the right panel of figure 19 demonstrates that symmetric pulses tend to be fainter and rarer than asymmetric pulses ; most of the symmetric pulses in the sample have low @xmath51 ratios .
overall , we find that eleven out of thirty isolated mask - weighted swift grb pulses are improved by the addition of the residual model .
these pulses exhibit complex shapes consistent with those detected in batse and fermi data , compared to 43 out of 50 batse grb pulses and two out of three fermi grb pulses .
when the potential model improvement resulting from the residual fits is considered as functions of @xmath51 ( figure 14 ) , @xmath41 ( figure 19 , left panel ) , and @xmath19 ( figure 19 , right panel ) , it appears that residual curves may be present in all isolated pulses , but are harder to identify in faint pulses .
it also appears that the residual features are harder to identify at swift s lower energies , even though swift and batse have similar instrumental sensitivities : _
a reasonable interpretation is that the residual features are less pronounced at lower energies . _
finally , residuals are easier to extract when pulses have large normalized residual peaks ( @xmath63 ) ; since symmetric pulses seem to have smaller @xmath41 values than asymmetric pulses , this suggests that there might be an observational bias favoring asymmetric pulses .
figures 20 through 25 summarize the hardness evolutions of the eleven swift grb pulses having measurable pulse shapes .
the swift hardness is defined to be the ratio of counts in the 50 to 300 kev energy range relative to the counts in the 15 to 50 kev energy range . in order to explore how
hardness changes with pulse shape , each hardness plot is marked with the five maxima and minima in the residual light curve : the precursor peak ( downward facing red arrow ) , the dip between the precursor peak and the central peak ( upward facing red arrow ) , the central plateau or peak ( downward facing black arrow ) , the dip between the central peak and the decay peak ( upward facing blue arrow ) , and the decay peak ( downward facing blue arrow ) .
these figures demonstrate the same general hard - to - soft evolution that characterizes batse and gbm grb pulses : the pulse rise phase is spectrally harder than the pulse decay phase .
furthermore , there seems to be a relationship between the times at which pulses harden and the three intensity peaks .
for example , two pulses ( grb 061004 and grb 080413b ) and possibly two others ( grb 070318 and grb 091018 ) have hardnesses that start soft and increase ( reaching maximum hardness at roughly the time of the central peak ) before subsequently decreasing in the normal fashion ; it is interesting to note that these three or four pulses are among the softest in the sample .
additionally , many of the pulses either undergo a brief re - hardening near the time of the decay peak : this is easily noticed for hard pulses grb 051111 , grb 060912a , and grb 061021 yet harder to discern for soft pulses .
table 4 summarizes the visual pulse structure as it relates to the maximum pulse hardness for swift pulses .
we wonder if similar coupled spectral and temporal behaviors exist for the previously - analyzed gbm and batse pulses , and therefore we have expanded the time periods of all plots from @xcite to correspond to the fiducial duration .
even though this redistribution of counts might result in low signal - to - noise , we recognize that the hardness evolution may be important over the entire life cycle of a grb pulses , and particularly in the time intervals corresponding to the residual wave features ( for this analysis , we note that the hardness ratio used batse data is the ratio of counts in the 100 kev to 1 mev energy range relative to the counts in the 25 to 100 kev energy range ) . individual pulse results for forty batse pulses having well - measured hardness evolutions
have been obtained for the entire fiducial timescale @xcite .
the hardness - related temporal structures for the two measurable gbm pulses are summarized in table 5 , while the 44 measurable batse pulse properties are shown in table 6 .
tables 4 , 5 , and 6 demonstrate interesting similarities between swift , gbm , and batse grb pulse evolutions : 1 ) hard pulses tend to have hard precursor peaks relative to the central peak , while soft pulses do not , and 2 ) hard pulses typically demonstrate a re - hardening of the decay peak , while soft pulses do not .
these results appear to be generally valid even though the energy ranges of the experiments are different .
thus , the complex pulse shapes uncovered for batse grb pulses are apparently ubiquitous , and span a larger energy range than just of that detected by batse .
we note the rather remarkable result that the hardness evolution seems to differ for hard pulses and soft pulses .
specifically , hard pulses seem more likely to evolve hard - to - soft while soft pulses are less likely to show this behavior .
furthermore , the times at which pulse evolution changes its decay rate appear linked to the times of the precursor peak , the central peak , and the decay peak . to explore in greater detail the hypothesis that pulse spectral evolution is associated with the residual structures , we examine the hardness evolution of batse pulses by grouping together those with similar maximum hardness ratios .
since the batse pulses in this sample all have different amplitudes , durations , and background count rates , it is difficult to average their time - dependent hardnesses in a reasonable yet unbiased way . the approach we choose is to sum together the 64 ms photon counts found in like parts of the fiducial window for similar pulses , then ratio these in different energy channels to obtain an evolving counts hardness ratio .
figure 26 indicates the resulting typical hardness evolution of the batse grb pulses , sorted into five groups using individual maximum hardness ratios .
hardnesses are taken by dividing the summed high energy counts ( 100 kev to 1 mev ) in each partition by the summed low energy counts ( 25 kev to 100 kev ) .
hardness groups are hr@xmath67 ( dark blue ) , @xmath68hr@xmath69 ( light blue ) , @xmath70hr@xmath71 ( dark green ) , @xmath72hr@xmath73 ( light green ) , and hr@xmath74 ( orange ) .
similar to the arrows shown in figures 20 through 25 , the down - facing arrows indicate the averaged fiducial positions of the precursor peak , the central peak , and the decay peak , while the up - facing arrows indicate the averaged fiducial positions of the gaps in the residual function between the peaks .
_ this figure demonstrates that all grb pulse groups evolve from hard to soft in that they are softer at the end of their fiducial time intervals than at the beginning ; in this sense all grb pulses evolve from hard to soft_. within this general hard - to - soft evolution , it is also true that grb pulse hardness is reinvigorated at the times of the three pulse peaks . in particular , the precursor peak seems to have a profound influence on the subsequent pulse evolution , since the pulses with the largest maximum hardness ratios have the most abrupt hard - to - soft evolutions and are hardest during the precursor peak , while those with the smallest maximum hardness ratios reach these values during the central peak .
there is weaker evidence in figure 25 for a decrease in the hardness decay during the decay peak than there is for the other two peaks .
@xcite have shown that hardness evolution is a function of pulse asymmetry : the short rise phases of asymmetric pulses cause them to have shorter lags than their symmetric counterparts , and lags are indicative of the rate of hardness decay . in this study
we have found that pulses with larger maximum hardnesses also decay faster , suggesting that maximum hardness should correlate with asymmetry .
figure 27 demonstrates that this correlation indeed exists and is in fact fairly robust : a spearman rank order test yields a p - value of only @xmath75 that these parameters are uncorrelated .
since asymmetry correlates with peak hardness , we expect to see that batse pulses grouped by asymmetry should exhibit hardness evolution that is also related to the timing of the three pulse peaks .
figure 28 demonstrates the effects of this grouping in a manner similar to figure 26 , but using @xmath19 as the key parameter rather than maximum hardness .
asymmetry groups are @xmath76 ( dark blue ) , @xmath77 ( light blue ) , @xmath78 ( dark green ) , @xmath79 ( light green ) , and @xmath80 ( orange ) .
the same general hard - to - soft evolution can be seen in the figure , although grouping the pulses by asymmetry may cause the precursor peak to be weaker than it is in figure 26 , but the rapid decay starting from the central peak is more pronounced .
maximum hardness is potentially a fragile classification characteristic , as is depends on how the hardnesses are binned in time ( e.g. bin size , quantity , and bin center time ) . to assure ourselves that the correlation of pulse evolution with hardness is robust , we group the pulses by hardness ( hr@xmath81 ) as measured from the ratio of the published energy fluences ( @xmath82 ,
where @xmath83 represents some combination of batse energy channels ) : @xmath84 the results , with hardness groups @xmath85 ( dark blue ) , @xmath86 ( light blue ) , @xmath87 ( dark green ) , @xmath88 ( light green ) , and @xmath89 ( orange ) , are shown in figure 29 .
these hardness groupings are entirely consistent with those shown in figure 26 .
the fact that they are also consistent with those shown in figure 28 demonstrates that grb pulse evolution has a similar functional dependence on both spectral hardness and asymmetry .
isolated swift grb pulses , like their batse and gbm counterparts , exhibit three - peaked pulse shapes even though swift observes the sky at lower energies than batse or gbm .
it is becoming increasingly clear that grb pulse structure , particularly at high signal - to - noise , can not be adequately modeled by simple monotonically increasing and decreasing functions , and spectrally - dependent variations must be treated as an integral part of each pulse .
isolated pulses generally evolve from hard to soft , although re - hardening often occurs coincident with each of the three peaks .
hard pulses are more likely to have a hard precursor peak and exhibit hard - to - soft evolution throughout .
soft pulses are more likely to have a soft precursor peak , causing them to reach maximum hardness at the time of the central pulse peak .
all isolated pulses are capable of having an additional hardening at the time of the decay peak ; this hardening is most often seen as a change in the rate of hardness decay .
_ these results strongly suggest that the commonly - used nomenclature of ` hard - to - soft ' and ` intensity tracking ' pulses is not correct ; all isolated grb pulses exhibit both behaviors simultaneously and to different degrees .
_ the re - hardening of grb pulse spectra coincident with the three temporal peaks suggests that a single - component evolving spectral model might be insufficient for producing characteristics of isolated grb pulses : it is reasonable to suspect that three separate but overlapping evolving components might exist .
@xcite have recently demonstrated through modeling that overlapping hard - to - soft pulses produce a composite hard - to - soft pulse exhibiting intensity tracking behaviors at the peaks .
if each of the three pulse components in isolated grb pulses evolves hard - to - soft , then our simple residual model is inadequate for describing these because it assumes that the first and third peaks are mirror images of one another , whereas the hardness evolution of these events would not be time - symmetric . more detailed , physical models are needed to understand the true nature of the underlying complex grb pulse structure . the standard model for grb prompt spectral emission is kinematic energy injection into a medium via collisionless relativistic shocks ; the preferred mechanism by which radiative energy is thought to be released is in the form of a synchrotron spectrum ( e.g. @xcite ) .
the physical internal shock model of @xcite , naturally accounts for three separate but temporally overlapping emission regions via a forward shock , a reverse shock , and the region containing the contact discontinuity , and this model may provide a physical basis for the underlying physical structures .
the model also predicts that the shock interactions overlap to a considerable extent , which can make them difficult to unambiguously observed separately .
kinematic models such as this one need to be merged with multi - component evolving spectral models , as recent multiple component spectral models ( e.g. @xcite ) , which thus far have not addressed the possibility that spectral evolution might be associated with specific light curve elements within an individual pulse .
@xcite have considered the radiation mechanisms of kinematically evolving synchrotron shocks , but find that the model better explains observed pulse spectral evolution if ( i ) the distribution of shock - accelerated electrons is steeper than generally assumed , ( ii ) the peak energy dependence is reduced by altering microphysics parameters , ( iii ) the lorentz factor undergoes more pronounced initial variations than previously thought , and ( iv ) the relativistic ejection proceeds with a constant mass flux rather than a constant kinetic energy flux .
@xcite indicate that these conditions are hard to justify , but that their ability to test microphysics conditions are limited by existing data .
further modeling is needed to see under what conditions the synchrotron shock model is consistent with observations .
single - episodic models of grb pulse shapes , such as those in which the asymmetric pulse shape is produced by curvature in a relativistic outflow ( e.g. @xcite ) predict neither the multi - episodic shapes nor the spectral rehardenings observed here .
these models likely need significant modifications in order to explain the observations .
complex grb pulses containing many more than three peaks may be simply the result of observing overlapping light curves from several isolated pulses . however , it is also possible that there are types of grb pulses different than those described here .
_ this latter hypothesis can not even be adequately explored without first having a working definition of a ` gamma - ray burst pulse ' such as has been summarized here for isolated grb pulses . _ the results obtained in this study
allow us to make general predications about the spectral evolution of grb - like pulses at other intrinsic energies .
for example , low - energy bursting events have been observed by transient experiments and include xrfs ( x - ray flashes ) and xrrs ( x - ray rich grbs ) : these appear to be low - energy counterparts to grbs ( e.g. @xcite ) .
similarly , the correlated behaviors of x - ray flares ( e.g. @xcite ) and optical flares ( e.g. @xcite ) observed in grb afterglows suggest that these are simply late , low - energy grb pulses . _
the correlated pulse behaviors identified in this manuscript suggest that low - energy pulses should exhibit time histories dominated by intensity tracking spectral behaviors , with a soft precursor peak and potentially a soft or non - existent decay peak .
_ we are currently exploring these possibilities in follow - up studies involving x - ray data and optical / infrared observations using etelman observatory at the university of the virgin islands @xcite .
in this paper and two prior ones @xcite we have re - conceived the observational definition of an isolated gamma - ray burst pulse .
we have been able to do this because isolated grb pulses have complex but well - defined and easily - replicated characteristics .
pulse spectral and temporal evolutions are intertwined in a way suggesting that the underlying physics is orderly and reproducible , even though titanic energies are unleashed through each grb in a relatively short period of time .
this study finds that swift grb pulses exhibit the same general characteristics as batse grb pulses : individual pulse light curves are characterized by triple - peaked profiles that generally evolve from hard to soft .
the three peaks ( the precursor peak , the central peak , and the decay peak ) are aligned to coincide with the asymmetry of the pulse shape .
the precursor and decay peaks can exhibit separate spectral evolution from that of the central peak , causing some pulses to follow intensity tracking evolution overlaying the natural hard - to - soft evolution .
the specific form that the evolution takes depends on the maximum spectral hardness of the pulse : hard pulses begin hard , and the overall hardness decays rapidly throughout the pulse .
however , during the decay , a re - hardening can occur as the decay peak intensity increases . pulses of intermediate hardness have precursor peaks having hardnesses similar to those of the central peak which decay at a moderate rate .
soft pulses can begin soft and harden at the central peak , after which the overall hardness decays slowly throughout the pulse .
the hardness of any pulse can remain constant or even increase during the time leading up to the decay peak .
_ as a result , ` hard - to - soft ' and ` intensity matching ' grb pulse definitions no longer appear to be valid , since all isolated pulses show both of these behaviors in ways that depend upon asymmetry and/or maximum hardness .
_ there is a strong interplay between a grb pulse s hardness and the symmetry of its temporal evolution : asymmetric pulses are generally hard and show a strong hard - to - soft evolution .
symmetric pulses are more often soft and exhibit intensity tracking spectral evolution .
a pulse s duration indicates the timescale over which a system is radiating ; this can be due to radiative cooling and/or to the timescale over which kinematic events occur .
a pulse s asymmetry and hard - to - soft evolution indicate that the rate of cooling is changing and/or that the kinematic events have more energy to release initially than at later times . the triple - peaked nature and spectral re - hardening of isolated grb pulses seems to indicate that energy is injected on as many as three separate occasions during the pulse duration .
theoretical models of grb prompt emission need to account for these observed pulse spectrotemporal characteristics .
we would like to acknowledge the helpful comments of the anonymous referee , as these greatly helped to improve the manuscript . this work has been supported by nasa epscor grant nnx13ad28a .
bonjak , . , & daigne , f. 2014 , , 568 , a45 broadbent , m. e. 2014 , ph.d .
burgess , j. m. , preece , r. d. , baring , m. g. , et al .
2011 , , 741 , 24 chincarini , g. , et al .
2010 , , 406 , 2113 crider , a. , liang , e. p. , preece , r. d. , briggs , m. s. , pendleton , g. n. , paciesas , w. s. , band , d. l. , & matteson , j. l. 1999 , , 138 , 401 dermer , c. d. 2004 , , 614 , 284 fenimore , e. e. , madras , c. d. , & nayakshin , s. 1996 , , 473 , 998 ford , l. a. , et al .
1995 , , 439 , 307 hakkila , j. , & preece , r. d. 2011 , , 740 , 104 hakkila , j. , & preece , r. d. 2014 , , 783 , 88 kaneko , y. , preece , r. d. , briggs , m. s. , paciesas , w. s. , meegan , c. a. , & band , d. l. 2006 , , 166 , 298 kino , m. , mizuta , a. , & yamada , s. 2004 , , 611 , 1021 li , l. , liang , e .- w . ,
tang , q .- w . , et al .
2012 , , 758 , 27 liang , e. , & kargatis , v. 1996 , , 381 , 49 lu , r .- j . ,
hou , s .- j . , &
liang , e .- w .
2010 , , 720 , 1146 margutti , r. , guidorzi , c. , chincarini , g. , bernardini , m. g. , genet , f. , mao , j. , & pasotti , f. 2010 , , 406 , 2149 markwardt , c. b. 2009 , astronomical data analysis software and systems xviii , 411 , 251 morris , d. c. , neff , j. e. , & hakkila , j. e. 2014 , american astronomical society meeting abstracts # 223 , 223 , # 148.42 norris , j. p. , bonnell , j. t. , kazanas , d. , scargle , j. d. , hakkila , j. , & giblin , t. w. 2005 , , 627 , 324 peng , z. y. , ma , l. , lu , r. j. , fang , l. m. , bao , y. y. , & yin , y. 2009 , , 14 , 311 peng , z. y. , zhao , x. h. , yin , y. , bao , y. y. , & ma , l. 2012 , , 752 , 132 preece , r. d. , goldstein , a. , bhat , n. , stanbro , m. , hakkila , j. & blalock , d. 2015 , submitted qin , y .- p .
2002 , , 396 , 705 ramirez - ruiz , e. , & fenimore , e. e. 2000 , , 539 , 712 rees , m. j. , & meszaros , p. 1994
, , 430 , l93 sakamoto , t. , lamb , d. q. , kawai , n. , et al .
2005 , , 629 , 311 scargle , j. d. 1998 , , 504 , 405 steiger , j. h. , shapiro , a. , & browne , m. b. 1985 , psychometrika , 50 , 253 .
willingale , r. , genet , f. , granot , j. , & obrien , p. t. 2010 , , 403 , 1296 grb 050801 & 5.8 & 4.1 & 0.999 & 0.100 & 1.46 & 0.09 & 0.08 & 0.12 & 0.21 & 1.35 & 2.92e-06 & 1.50e-02 & 1.94e+00 & 2.12e-01 & 1.00 &
1556 + grb 050908 & 19.8 & 56.0 & 0.053 & 0.286 & 0.70 & 0.09 & -77.87 & 134.36 & 0.06 & 0.01 & 8.15e+03 & 4.10e+04 & 7.91e-01 & 1.32e+00 & 1.01 & 6233 + grb 051111 & 43.4 & 1.9 & 0.708 & 0.032 & 2.66 & 0.13 & -7.88 & 0.36 & 0.28 & 0.01 & 5.69e+00 & 1.03e+00 & 1.02e+01 & 4.91e-01 & 1.02 & 6243 + grb 060206 & 8.3 & 1.4 & 0.209 & 0.064 & 2.79 & 0.10 & -7.18 & 1.95 & 0.35 & 0.01 & 1.55e+02 & 9.76e+01 & 5.75e-01 & 1.19e-01 & 1.01 & 6243 + grb 060512 & 18.3 & 111.3 & 0.032 & 0.373 & 0.88 & 0.12 & -140.92 & 1103.85 & 0.06 & 0.01 & 1.06e+05 & 2.44e+06 & 1.94e-01 & 1.50e+00 & 0.95 & 697 + grb 060708 & 9.2 & 1.2 & 0.518 & 0.099 & 1.94 & 0.09 & -1.85 & 0.51 & 0.18 & 0.01 & 6.63e+00 & 3.43e+00 & 1.58e+00 & 2.54e-01 & 1.11 & 774 + grb 060912a & 5.1 & 0.3 & 0.750 & 0.042 & 8.58 & 0.27 & -0.44 & 0.05 & 0.97 & 0.03 & 4.28e-01 & 1.09e-01 & 1.26e+00 & 7.68e-02 & 1.08 & 774 + grb 060926 & 9.7 & 1.5 & 0.840 & 0.094 & 1.09 & 0.09 & -0.25 & 0.13 & 0.13 & 0.01 & 2.63e-01 & 2.00e-01 & 2.71e+00 & 4.55e-01 & 1.06 & 618 + grb 061004 & 7.1 & 2.3 & 0.161 & 0.099 & 2.54 & 0.10 & -7.49 & 4.45 & 0.26 & 0.01 & 3.04e+02 & 3.79e+02 & 3.83e-01 & 1.58e-01 & 1.04 & 6243 + grb 061021 & 11.5 & 0.4 & 0.844 & 0.024 & 6.11 & 0.16 & 1.66 & 0.04 & 0.66 & 0.01 & 2.95e-01 & 5.85e-02 & 3.23e+00 & 1.19e-01 & 1.17 & 3118 + grb 070318 & 51.1 & 2.2 & 0.921 & 0.026 & 1.76 & 0.09 & -0.83 & 0.10 & 0.17 & 0.00 & 2.84e-01 & 1.04e-01 & 1.57e+01 & 7.14e-01 & 1.01 & 4681 + grb 070506 & 9.2 & 10.5 & 0.114 & 0.245 & 0.96 & 0.08 & -13.75 & 28.46 & 0.09 & 0.01 & 1.14e+03 & 4.87e+03 & 3.52e-01 & 5.02e-01 & 1.00 & 3899 + grb 070611 & 18.6 & 327.4 & 0.012 & 0.413 & 0.82 & 0.13 & -377.70 & 8441.82 & 0.06 & 0.01 & 1.90e+06 & 1.26e+08 & 7.60e-02 & 1.69e+00 & 1.04 & 696 + grb 070810a & 13.1 & 2.3 & 0.444 & 0.119 & 1.90 & 0.12 & -5.10 & 1.24 & 0.18 & 0.01 & 1.82e+01 & 1.20e+01 & 1.94e+00 & 4.04e-01 & 1.01 & 9986 + grb 071010b & 18.3 & 0.4 & 0.624 & 0.017 & 7.70 & 0.18 & -2.40 & 0.11 & 0.85 & 0.01 & 5.28e+00 & 4.63e-01 & 3.81e+00 & 9.47e-02 & 1.08 & 4680 + grb 071031 & 44.7 & 12.2 & 0.585 & 0.204 & 0.50 & 0.06 & -10.89 & 3.87 & 0.04 & 0.00 & 1.82e+01 & 1.90e+01 & 8.70e+00 & 2.79e+00 & 1.00 & 3119 + grb 071117 & 5.4 & 0.2 & 0.659 & 0.029 & 11.3 & 0.4 & -0.80 & 0.05 & 1.44 & 0.02 & 1.13e+00 & 1.72e-01 & 1.19e+00 & 4.84e-02 & 1.10 & 462 + grb 080129 & & & & & 0.50 & 0.12 + grb 080413b & 4.8 & 0.3 & 0.450 & 0.040 & 18.70 & 0.49 & -1.94 & 0.15 & 2.07 & 0.04 & 6.33e+00 & 1.39e+00 & 7.25e-01 & 4.90e-02 & 1.03 & 3274 + grb 080430 & 14.8 & 0.8 & 0.725 & 0.040 & 2.60 & 0.12 & -0.94 & 0.14 & 0.32 & 0.01 & 1.65e+00 & 3.80e-01 & 3.58e+00 & 2.18e-01 & 1.05 & 1087 + grb 080710 & 23.0 & 5.0 & 0.853 & 0.132 & 1.00 & 0.12 & -0.70 & 0.39 & 0.08 & 0.01 & 5.13e-01 & 5.81e-01 & 6.56e+00 & 1.50e+00 & 0.99 & 2337 + grb 080804 & 36.4 & 2.8 & 0.781 & 0.060 & 3.10 & 0.24 & -1.64 & 0.41 & 0.28 & 0.01 & 2.17e+00 & 8.45e-01 & 9.47e+00 & 8.12e-01 & 0.98 & 2337 + grb 080805 & 74.4 & 5.4 & 0.787 & 0.043 & 1.10 & 0.06 & -5.62 & 0.57 & 0.11 & 0.00 & 4.16e+00 & 1.18e+00 & 1.95e+01 & 1.57e+00 & 1.02 & 2337 + grb 081007 & 9.9 & 1.5 & 0.672 & 0.117 & 2.60 & 0.24 & -0.69 & 0.35 & 0.27 & 0.02 & 1.85e+00 & 1.17e+00 & 2.23e+00 & 3.84e-01 & 1.01 & 2337 + grb 081118 & 98.4 & 41.0 & 0.276 & 0.198 & 0.60 & 0.12 & -38.57 & 40.86 & 0.04 & 0.00 & 7.52e+02 & 1.15e+03 & 9.05e+00 & 4.70e+00 & 0.99 & 3274 + grb 081222 & 15.6 & 0.5 & 0.453 & 0.023 & 7.70 & 0.12 & -3.23 & 0.26 & 0.94 & 0.01 & 1.98e+01 & 2.48e+00 & 2.35e+00 & 9.24e-02 & 1.51 & 1087 + grb 090313 & & & & & 0.80 & 0.18 + grb 090809a & 11.9 & 11.1 & 0.191 & 0.331 & 1.10 & 0.12 & -12.14 & 17.55 & 0.10 & 0.01 & 2.97e+02 & 1.05e+03 & 7.54e-01 & 8.86e-01 & 0.97 & 696 + grb 090927 & 3.0 & 0.5 & 0.920 & 0.131 & 2.00 & 0.12 & 0.01 & 0.02 & 0.31 & 0.04 & 1.71e-02 & 3.18e-02 & 9.35e-01 & 1.57e-01 & 1.08 & 2337 + grb 091018 & 5.4 & 0.3 & 0.399 & 0.034 & 10.30 & 0.24 & -1.55 & 0.18 & 1.14 & 0.02 & 1.11e+01 & 2.24e+00 & 7.13e-01 & 4.52e-02 & 1.03 & 2336 + columns : ( 1 ) burst i d , ( 2 ) pulse duration @xmath14 , ( 3 ) duration uncertainty @xmath90 , ( 4 ) asymmetry @xmath19 , ( 5 ) asymmetry uncertainty @xmath91 , ( 6 ) peak flux @xmath92 ( @xmath93 ) , ( 7 ) peak flux uncertainty @xmath94 ( @xmath93 ) , ( 8) start time @xmath8 ( s ) , ( 9 ) start time uncertainty @xmath95 ( s ) , ( 10 ) amplitude @xmath7 , ( 11 ) amplitude uncertainty @xmath96 ( s ) , ( 12 ) rise parameter @xmath9 , ( 13 ) rise parameter uncertainty @xmath97 , ( 14 ) decay parameter @xmath10 , ( 15 ) decay parameter uncertainty @xmath98 , ( 16 ) @xmath99 per degree of freedom @xmath100 , and ( 17 ) degrees of freedom @xmath100 . grb080129 and grb 090313 could not be fitted .
lrrrrrrcc grb pulse & @xmath27 & @xmath25 & @xmath101 & @xmath102 & @xmath103 & @xmath104 & p for @xmath44 & significant ? + grb 050801 & -1.501 & 15.890 & 1.00 & 267 & 1.00 & 263 & @xmath105 & + grb 050908 & -80.452 & 28.326 & 0.98 & 1695 & 0.97 & 1691 & @xmath106 & + grb 051111 & -16.495 & 84.408 & 1.04 & 1572 & 1.01 & 1568 & @xmath107 & yes + grb 060206 & -8.108 & 11.548 & 1.14 & 303 & 1.07 & 299 & @xmath108 & yes + grb 060512 & -142.254 & 15.719 & 0.95 & 2504 & 0.95 & 2500 & @xmath109 & + grb 060708 & -3.384 & 16.702 & 1.34 & 308 & 1.17 & 304 & @xmath110 & yes + grb 060912a & -1.519 & 11.104 & 1.17 & 187 & 0.93 & 183 & @xmath111 & yes + grb 060926 & -2.495 & 23.078 & 1.12 & 394 & 1.12 & 390 & 1.0 & + grb 061004 & -8.235 & 10.784 & 1.58 & 235 & 1.21 & 231 & @xmath112 & yes + grb 061021 & -0.760 & 26.887 & 1.87 & 427 & 1.44 & 423 & @xmath113 & yes + grb 070318 & -13.591 & 128.84 & 1.08 & 2221 & 1.06 & 2217 & @xmath114 & yes + grb 070506 & -14.649 & 15.275 & 1.02 & 463 & 1.02 & 459 & @xmath115 & + grb 070611 & -379.247 & 17.260 & 1.00 & 4013 & 1.00 & 4009 & 1.0 & + grb 070810a & -7.116 & 20.954 & 0.99 & 463 & 0.98 & 459 & @xmath116 & + grb 071010b & -5.841 & 36.544 & 1.19 & 621 & 1.18 & 617 & @xmath117 & + grb 071031 & -18.935 & 82.172 & 1.00 & 1546 & 1.00 & 1542 & @xmath118 & + grb 071117 & -1.857 & 10.901 & 1.36 & 194 & 1.31 & 190 & @xmath119 & yes + grb 080129 & & & & & & & & + grb 080413b & -2.668 & 7.669 & 1.94 & 156 & 1.45 & 152 & @xmath120 & yes + grb 080430 & -4.023 & 32.346 & 1.05 & 563 & 1.05 & 559 & @xmath121 & + grb 080710 & -6.119 & 55.364 & 1.03 & 956 & 1.03 & 952 & @xmath122 & + grb 080804 & -9.643 & 82.956 & 1.01 & 1442 & 1.00 & 1438 & @xmath123 & + grb 080805 & -22.089 & 168.045 & 1.06 & 2965 & 1.05 & 2961 & @xmath124 & + grb 081007 & -2.651 & 20.996 & 1.01 & 365 & 1.00 & 361 & @xmath125 & + grb 081118 & -46.324 & 121.500 & 1.00 & 3372 & 1.00 & 3368 & 1.0 & + grb 081222 & -5.638 & 27.706 & 2.14 & 516 & 1.85 & 512 & @xmath126 & yes + grb 090313 & & & & & & & & + grb 090809a & -13.438 & 15.806 & 1.01 & 452 & 1.01 & 450 & 1.0 & + grb 090927 & -0.750 & 7.742 & 1.10 & 128 & 1.04 & 124 & @xmath127 & + grb 091018 & -2.324 & 9.042 & 1.49 & 172 & 1.15 & 168 & @xmath110 & yes + lrrrrrrrrrrrrr grb pulse & @xmath30 & @xmath128 & @xmath31 & @xmath129 & @xmath32 & @xmath130 & @xmath33 & @xmath131 & r + grb 051111 & -0.01 & 0.56 & 0.053 & 0.003 & 9.5 & 0.9 & 0.18 & 0.02 & 0.19 + grb 060206 & 2.93 & 0.12 & 0.047 & 0.004 & 9.7 & 0.8 & 1.00 & 0.13 & 0.13
+ grb 060708 & 1.55 & 0.04 & 0.097 & 0.005 & 22.1 & 1.4 & 0.53 & 0.04 & 0.55 + grb 060912a & 0.33 & 0.02 & 0.189 & 0.010 & 73.0 & 10.3 & 0.15 & 0.02 & 0.20 + grb 061004 & 2.97 & 0.05 & 0.098 & 0.004 & 21.9 & 1.4 & 0.53 & 0.04 & 0.38 + grb 061021 & 2.78 & 0.04 & 0.186 & 0.007 & 32.2 & 2.4 & 0.19 & 0.02 & 0.28 + grb 070318 & 2.52 & 0.18 & 0.031 & 0.002 & 11.9 & 2.0 & 0.11 & 0.02 & 0.19 + grb 071117 & 0.61 & 0.10 & 0.090 & 0.020 & 54.8 & 16.3 & 0.70 & 0.29 & 0.06 + grb 080413b & 0.40 & 0.02 & 0.423 & 0.020 & 35.8 & 2.5 & 0.37 & 0.03 & 0.20 + grb 081222 & 4.10 & 0.08 & 0.113 & 0.005 & 18.0 & 1.7 & 0.20 & 0.02 & 0.12 + grb 091018 & 1.02 & 0.04 & 0.146 & 0.008 & 35.6 & 3.1 & 0.31 & 0.035 & 0.13 + lccc swift pulse & max cts hardness & precursor peak & decay peak re - hardening ? + grb 081222 & @xmath132 & hard & yes + grb 071117 & @xmath133 & hard & yes + grb 051111 & @xmath134 & hard & yes + grb 061021 & @xmath135 & moderate & maybe + grb 060912a & @xmath136 & hard & yes + grb 060708 & @xmath137 & moderate & yes + grb 080413b & @xmath138 & soft & maybe + grb 070318 & @xmath139 & moderate & no + grb060206 & @xmath140 & hard & maybe + grb 061004 & @xmath141 & soft & no + grb 091018 & @xmath142 & moderate & maybe + lccc batse pulse & max cts hardness & precursor peak & decay peak re - hardening ? + batse 3040 & @xmath145 & hard & yes + batse 1883 & @xmath146 & hard & yes + batse 563 & @xmath147 & moderate & maybe + batse 907 & @xmath148 & moderate & maybe + batse 3257 & @xmath149 & hard & no + batse 7711 & @xmath150 & hard & maybe + batse 1406 & @xmath151 & hard & maybe + batse 469 & @xmath152 & moderate & yes + batse 2662 & @xmath153 & moderate & maybe + batse 2665 & @xmath154 & hard & yes + batse 3143 & @xmath155 & hard & maybe + batse 3003 & @xmath156 & hard & maybe + batse 658 & @xmath157 & hard & yes + batse 3164 & @xmath158 & soft & yes + batse 1580 & @xmath159 & moderate & maybe + batse 1446 & @xmath160 & soft & maybe + batse 332 & @xmath161 & hard & yes + batse 1467 & @xmath162 & hard & maybe + batse 1806 & @xmath163 & hard & yes + batse 7614 & @xmath164 & moderate & no + batse 501 & @xmath165 & moderate & no + batse 2958p1 & @xmath166 & hard & maybe + batse 7903 & @xmath167 & moderate & yes + batse 7638 & @xmath168 & soft & yes + batse 7843 & @xmath169 & soft & maybe + batse 7989 & @xmath170 & soft & yes + batse 3168 & @xmath171 & soft & no + batse 914 & @xmath172 & soft & no + batse 1145 & @xmath173 & moderate & maybe + batse 1306 & @xmath174 & moderate & no + batse 2958p2 & @xmath175 & soft & no + batse 680 & @xmath176 & soft & maybe + batse 795 & @xmath177 & soft & no + batse 1039 & @xmath178 & moderate & no + batse 540 & @xmath179 & moderate & maybe + batse 2862 & @xmath180 & soft & maybe + batse 3026 & @xmath181 & moderate & maybe + batse 8112 & @xmath182 & soft & no + batse 493 & @xmath141 & soft & no + batse 1432 & @xmath183 & moderate & maybe + batse 1200 & @xmath184 & moderate & maybe + batse 8121 & @xmath185 & soft & no + batse 1319 & @xmath186 & soft & no + batse 1379 & @xmath187 & soft & no + $ ] ) is shown as a function of the signal - to - noise ratio ( @xmath51 ) .
a large @xmath188 $ ] value indicates improvement consistent with a better model fit , while a small one suggests only slight improvement .
batse fits ( crosses ) and fermi pulses ( open diamonds ) improve systematically with increasing signal - to - noise , suggesting that complex pulse shapes are normal at high signal - to - noise but get harder to recognize at lower signal - to - noise .
swift pulses ( filled circles ) have similar @xmath51 ratios to batse and fermi pulses . [ fig14 ] ] vs. @xmath19 for the pulses in this analysis ; there is no obvious correlation .
swift pulses are indicated by filled circles , batse pulses by crosses , and fermi pulses by diamonds . *
the horizontal line shows the mean value @xmath189 * [ fig18 ] ] ( dark blue ) , eight have @xmath68hr@xmath69 ( light blue ) , nine have @xmath190hr@xmath71 ( dark green ) , ten have @xmath72hr@xmath73 ( light green ) , and ten have hr@xmath74 ( orange ) .
the effect of the precursor peak is clearly important in defining whether a pulse undergoes hard - to - soft or intensity matching evolution .
[ fig26 ] ] .
individual pulses are partitioned using residual structures ( arrows are as described in figure 26 ) , then the counts in each partition are summed for all pulses having similar fluence hardnesses .
hardnesses are obtained for these binned groups as described for figure 26 .
eight pulses have hr@xmath193 ( dark blue ) , nine have @xmath194hr@xmath195 ( light blue ) , eight have @xmath196hr@xmath197 ( dark green ) , eight have @xmath198hr@xmath199 ( light green ) , and eleven have @xmath200hr@xmath201 ( orange ) . [ fig29 ] ] | isolated swift gamma - ray burst ( grb ) pulses , like their higher - energy batse counterparts , emit the bulk of their pulsed emission as a hard - to - soft component that can be fitted by the @xcite empirical pulse model .
this signal is overlaid by a fainter , three - peaked signal that can be modeled by an empirical wave - like function @xcite : the two fits combine to reproduce grb pulses with distinctive three - peaked shapes .
the precursor peak appears on or before the pulse rise and is often the hardest component , the central peak is the brightest , and the decay peak converts exponentially decaying emission into a long , soft , power - law tail .
accounting for systematic instrumental differences , the general characteristics of the fitted pulses are remarkably similar .
isolated grb pulses are dominated by hard - to - soft evolution ; this is more pronounced for asymmetric pulses than for symmetric ones .
isolated grb pulses can also exhibit intensity tracking behaviors that , when observed , are tied to the timing of the three peaks : pulses with the largest maximum hardnesses are hardest during the precursor , those with smaller maximum hardnesses are hardest during the central peak , and all pulses can re - harden during the central peak and/or during the decay peak . since these behaviors are essentially seen in all isolated pulses , the distinction between `` hard - to - soft '' and `` intensity - tracking '' pulses really no longer applies .
additionally , the triple - peaked nature of isolated grb pulses seems to indicate that energy is injected on three separate occasions during the pulse duration : theoretical pulse models need to account for this . |
from pdg2006@xcite , the @xmath8 meson nonet ( @xmath9 , @xmath10 , @xmath11 , and @xmath12 ) as well as the @xmath13 members [ @xmath14 , @xmath15 , and @xmath16 has been well established . in ref.@xcite
, we suggested that the @xmath17 , @xmath18 , together with the @xmath19 and @xmath20 observed by bes collaboration@xcite constitute the @xmath21 meson nonet .
more recently , we argued that the @xmath4 with a mass of @xmath22 mev and a width of @xmath23 mev observed by the bes collaboraton@xcite could be the @xmath5 @xmath24 in ref .
@xcite where the other members of the @xmath5 meson nonet were not discussed . in the present work , we shall address the possible assignment for the @xmath5 meson nonet . with the assignment of the @xmath4 as the @xmath24 member of the @xmath5 meson nonet
, one can expect that both the isovector and another isoscalar members of the @xmath5 meson nonet should be lighter than the @xmath4 .
experimentally , in the mass region 2000 - 2225 mev , the pseudoscalar states @xmath0 [ mass : @xmath25 mev , width : @xmath26 mev ] , @xmath1 [ mass : @xmath27 mev , width : @xmath28 mev ] , @xmath2 [ mass : @xmath29 mev , width : @xmath30 mev ] , and @xmath3 [ mass : @xmath31 mev , width : @xmath32 mev ] are reported@xcite . theoretically , some predicted values for the @xmath33 mass are 2.15 gev by qcd sum rules@xcite , 2.009 gev by the spectrum integral equation@xcite , 2.193 gev by a covariant quark model@xcite , 2.039 gev by a relativistic independent quark model@xcite , and 2.07 gev by regge phenomenology@xcite .
in addition , the mass of the third radial excitation of the @xmath10 is predicted to be about 2.267 gev by a covariant quark model@xcite or 2.1 gev by regge phenomenology@xcite .
the @xmath0 mass is similar to the predicted @xmath33 mass , and all the masses of the @xmath1 , @xmath2 , and @xmath3 are close to the predicted mass range of the third radial excitation of the @xmath10 .
only the mass information of these states is insufficient to classify them .
the main purpose of this work is to discuss whether these reported pseudoscalar states can be assigned as the members of the @xmath5 meson nonet or not by investigating their decay properties in the @xmath6 meson decay model .
the organization of this paper is as follows . in section 2 ,
the brief review of the @xmath6 decay model is given ( for the detailed review see _
e.g. _ refs.@xcite . ) in sections 3 and 4 , the decay widths of the @xmath0 , @xmath1 , @xmath2 , @xmath3 , and @xmath4 as the @xmath5 @xmath7 state are presented .
the decay widths of the @xmath5 kaon are predicted in section 5 , and the summary and conclusion are given in section 6 .
the @xmath6 decay model , also known as the quark - pair creation model , was originally introduced by micu@xcite and further developed by le yaouanc et al.@xcite . the @xmath6 decay model which ( in several variants ) is the standard model for strong decays at least for mesons in the initial state , has been widely used to evaluate the strong decays of hadrons@xcite , since it gives a good description of many of the observed decay amplitudes and partial widths of the hadrons . the main assumption of the @xmath6 decay model is that strong decays take place via the creation of a @xmath6 quark - antiquark pair from the vacuum . the new produced quark - antiquark pair , together with the @xmath7 within the initial meson regroups into two outgoing mesons in all possible quark rearrangement ways , which corresponds to the two decay diagrams as shown in fig.1 for the meson decay process @xmath34 . the transition operator @xmath35 of the decay @xmath36 in the @xmath6 model is given by @xmath37 where @xmath38 is a dimensionless parameter representing the probability of the quark - antiquark pair @xmath39 with @xmath40 creation from the vacuum , @xmath41 and @xmath42 are the momenta of the created quark @xmath43 and antiquark @xmath44 , respectively .
@xmath45 , @xmath46 , and @xmath47 are the flavor , color , and spin wave functions of the @xmath39 , respectively .
the solid harmonic polynomial @xmath48 reflects the momentum - space distribution of the @xmath39 . for the meson wave function ,
we adopt the mock meson @xmath49 defined by@xcite @xmath50 where @xmath51 and @xmath52 are the masses of the quark @xmath53 with a momentum of @xmath54 and the antiquark @xmath55 with a momentum of @xmath56 , respectively .
@xmath57 is the radial quantum number of the meson @xmath58 composed of @xmath59 .
@xmath60 , @xmath61 , @xmath62 ( @xmath63 ) is the spin of @xmath53 ( @xmath64 ) , @xmath65 is the relative orbital angular momentum between @xmath53 and @xmath64 .
@xmath66 , @xmath67 .
@xmath68 is a clebsch - gordan coefficient , and @xmath69 is the total energy of the meson @xmath58 .
@xmath70 , @xmath71 , @xmath72 , and @xmath73 are the spin , flavor , color , and space wave functions of the meson @xmath58 , respectively .
the mock meson satisfies the normalization condition @xmath74 the @xmath75-matrix of the process @xmath36 is defined by @xmath76 with @xmath77 where @xmath78 is the helicity amplitude of @xmath36 . in the center of mass frame of meson @xmath58 ,
@xmath78 can be written as @xmath79{l } \scriptstyle m_{l_a},m_{s_a},\\\scriptstyle m_{l_b},m_{s_b},\\ \scriptstyle m_{l_c},m_{s_c},m \end{array}}\renewcommand{\arraystretch}{1}\!\ !
\langle l_am_{l_a}s_am_{s_a}|j_am_{j_a}\rangle\nonumber\\ & & \times\langle l_bm_{l_b}s_bm_{s_b}|j_bm_{j_b}\rangle\langle l_cm_{l_c}s_cm_{s_c}|j_cm_{j_c}\rangle\nonumber\\ & & \times\langle 1m1-m|00\rangle\langle \chi^{14}_{s_bm_{s_b}}\chi^{32}_{s_cm_{s_c}}|\chi^{12}_{s_am_{s_a}}\chi^{34}_{1-m}\rangle\nonumber\\ & & \times[f_1i(\vec{p},m_1,m_2,m_3)+(-1)^{1+s_a+s_b+s_c}f_2i(-\vec{p},m_2,m_1,m_3)],\end{aligned}\ ] ] with @xmath80 and @xmath81 , corresponding to the contributions from figs . 1 ( a ) and 1 ( b ) , respectively , and @xmath82 where @xmath83 , @xmath84
, @xmath85 is the mass of the created quark @xmath43 .
the spin overlap in terms of winger s @xmath86 symbol can be given by @xmath87 in order to compare with the experiment conventionally , @xmath88 can be converted into the partial amplitude by a recoupling calculation@xcite @xmath89{l } \scriptstyle m_{j_b},m_{j_c},\\\scriptstyle m_s , m_l \end{array}}\renewcommand{\arraystretch}{1}\!\ ! \langle lm_lsm_s|j_am_{j_a}\rangle\langle j_bm_{j_b}j_cm_{j_c}|sm_s\rangle\nonumber\\ & & \times\int d\omega\,\mbox{}y^\ast_{lm_l}{\cal{m}}^{m_{j_a}m_{j_b}m_{j_c } } ( \vec{p}).\end{aligned}\ ] ] if we consider the relativistic phase space , the decay width @xmath90 in terms of the partial wave amplitudes is @xmath91 here @xmath92=@xmath93[m^2_a-(m_b - m_c)^2]}}{2m_a}$ ] , @xmath94 , @xmath95 , and @xmath96 are the masses of the meson @xmath58 , @xmath97 , and @xmath98 , respectively
. the decay width can be derived analytically if the simple harmonic oscillator ( sho ) approximation for the meson space wave functions is used . in momentum - space ,
the sho wave function is @xmath99 where the radial wave function is given by @xmath100 here @xmath101 is the sho wave function scale parameter , and @xmath102 is an associated laguerre polynomial .
the sho wave functions can not be regarded as realistic , however , they are a _ de facto _ standard for many nonrelativistic quark model calculations . moreover
, the more realistic space wave functions such as those obtained from coulomb , plus the linear potential model do not always result in systematic improvements due to the inherent uncertainties of the @xmath6 decay model itself@xcite .
the sho wave function approximation is commonly employed in the @xmath6 decay model in literature . in the present work , the sho wave function approximation for the meson space wave functions is taken . under the sho wave function approximation ,
the parameters used in the @xmath6 decay model involve the @xmath7 pair production strength parameter @xmath38 , the sho wave function scale parameter @xmath101 , and the masses of the constituent quarks .
in the present work , we take @xmath103 and @xmath104 gev , the values recently obtained by fitting 32 experimentally well - determined decay rates with the @xmath6 decay model is higher than that used by ref.@xcite ( 0.505 ) by a factor of @xmath105 due to different field conventions , constant factor in @xmath35 , etc
. the calculated results of the widths are , of course , unaffected . ] , and @xmath106 gev , @xmath107 gev@xcite .
the meson masses used to determine the phase space and final state momenta are as the @xmath108 kaon is problematic@xcite .
quark model@xcite and other phenomenological approaches@xcite consistently suggest the @xmath108 kaon has a mass about 1580 mev , here we take 1580 mev as the mass of the @xmath108 kaon [ @xmath109 . also , we assume that the @xmath110 , @xmath111 , and @xmath112 are the ground scalar mesons as refs.@xcite . ]
@xmath113 mev , @xmath114 mev , @xmath115 mev , @xmath116 mev , @xmath117 mev , @xmath118 mev , @xmath119 mev , @xmath120 mev , @xmath121 mev , @xmath122 mev , @xmath123 mev , @xmath124 mev , @xmath125 mev , @xmath126 mev , @xmath127 mev , @xmath128 mev , @xmath129 mev , @xmath130 mev , @xmath131 mev , @xmath132 mev , @xmath133 mev , and @xmath134 mev .
from ( [ width1 ] ) , the numerical values of the partial decay widths of the @xmath0 as the @xmath5 isovector state are listed in table 1 .
the initial state mass is set to @xmath135 mev .
cc|ccmode & @xmath136 ( mev ) & mode & @xmath136 ( mev ) + @xmath137 & 3.1 & @xmath138 & 5.9 + @xmath139 & 52.0 & @xmath140 & 3.6 + @xmath141 & 112.5 & @xmath142 & 48.0 + @xmath143 & 0.3 &
@xmath144 & 8.8 + @xmath145 & 5.1 & @xmath146 & 8.5 + @xmath147 & 14.9 & @xmath148 & 10.3 + @xmath149 & 4.6 & + + table 1 indicates that the total width of the @xmath0 as the @xmath5 isovector state predicted by the @xmath6 decay model is about @xmath150 mev , consistent with the observation @xmath151 mev within errors , and the dominant decay modes are expected to be @xmath139 , @xmath141 and @xmath152 .
also , in order to check the dependence of the theoretical result on the initial state mass , the predicted total width of the @xmath0 is shown in fig .
2 as the function of the initial state mass . fig .
2 shows that when the initial state mass varies from 2035 to 2105 mev , the total width of the @xmath5 isovector state varies from about 230 to 320 mev , generally in accord with the width range of the @xmath0 .
both the mass and width of the @xmath0 are consistent with the predicted @xmath5 isovector state , which therefore suggests that the assignment of the @xmath0 as the @xmath5 isovector state seams plausible .
in the presence of the @xmath4 being one isoscalar member of the @xmath5 meson nonet , the @xmath1 , @xmath2 , and @xmath3 would complete another isoscalar member .
it is well known that in a meson nonet , the two physical isoscalar states can mix .
the mixing of the two isoscalar states can be parameterized as @xmath153 where @xmath154 and @xmath24 are the pure @xmath5 nonstrange and strange states , respectively , and @xmath155 denotes the @xmath1 , @xmath2 , or @xmath3 . according to ( [ width1 ] ) ,
the partial widths of @xmath155 and @xmath4 become with mixing @xmath156 based on ( [ w1 ] ) and ( [ w2 ] ) , the predicted total widths of the @xmath1 , @xmath2 , @xmath3 , and @xmath4 are shown in fig .
3 as functions of the initial state mass and the mixing angle @xmath157 . from fig .
3 , one can see that with the variations of the initial state mass and @xmath157 , only the measured widths of the @xmath2 and @xmath4 are possible to be reasonably reproduced in the @xmath6 model .
we therefore suggest that the assignment of the @xmath1 and @xmath158 as the @xmath5 isoscalar states seems unfavorable .
we shall focus on the possibility of the @xmath2 being the partner of the @xmath4 . taking @xmath159 mev and @xmath160 mev , we list the numerical values of the partial decay widths of the @xmath2 and @xmath4 in table 2 .
the variation of the theoretical total widths of the @xmath2 and @xmath4 with the mixing angle @xmath157 is shown in fig .
.decays of the @xmath2 and @xmath4 as the @xmath5 isoscalar states in the @xmath6 model .
@xmath161 , @xmath162 .
[ cols="^,^,^ " , ] from fig .
4 , we find that if the @xmath163 mixing angle @xmath157 lying in the range from about @xmath164 to @xmath165 radians , both the measured widths of the @xmath2 and @xmath4 can be reasonably reproduced . in order to check whether the possibility of @xmath166 radians exists or not , below we shall estimate the @xmath2-@xmath4 mixing angle phenomenologically . in the @xmath167 and @xmath24 basis , the mass - squared matrix describing the @xmath2 and @xmath4 mixing can be written as@xcite @xmath168 where @xmath169 and @xmath170 are the masses of the states @xmath167 and @xmath24 , respectively , @xmath171 denotes the total annihilation strength of the @xmath7 pair for the light flavors @xmath172 and @xmath173 , @xmath174 describes the @xmath175-breaking ratio of the nonstrange and strange quark masses via the constituent quark mass ratio @xmath176 .
the masses of the two physical states @xmath2 and @xmath4 can be related to the matrix @xmath177 by the unitary matrix @xmath178 @xmath179 @xmath167 is the orthogonal partner of the @xmath180 , the isovector state of @xmath5 meson nonet , and one can expect that @xmath167 degenerates with @xmath33 in effective quark masses , here we take @xmath181 . with the help of the gell - mann - okubo mass formula @xmath182@xcite , the following relations can be derived from ( [ diag ] ) @xmath183\nonumber\\ & & \times[(2-x^2)m^2_{\pi(2070)}+(2+x^2)m^2_{\eta(2225)}-4m^2_{k(4 ^ 1s_0 ) } ] , \label{msch}\end{aligned}\ ] ] @xmath184 if the @xmath175-breaking effect is not considered , i.e. , @xmath185 , relation ( [ msch ] ) can be reduced to schwinger s original nonet mass formula@xcite . taking @xmath186 , from ( [ msch ] ) and ( [ am ] )
we have @xmath187 based on the values of the above parameters involved in ( [ matrix ] ) , the unitary matrix @xmath188 can be given by @xmath189 which gives @xmath190 radians , just lying in the range from about @xmath164 to @xmath165 radians . from table 2 , this estimated mixing angle leads to @xmath191 mev and @xmath192 mev , both in good agreement with the experimental results .
the @xmath2 and @xmath4 , together with the @xmath0 therefore appear to be the convincing @xmath5 states .
the above two sections show that in the presence of the @xmath0 , @xmath2 , and @xmath4 belonging to the @xmath5 meson nonet , the total widths of these three states can be naturally accounted for in the @xmath6 decay model , and the @xmath5 kaon is expected to have a mass of about 2153 mev by the mass formula ( [ msch ] ) . below the @xmath193
denotes the @xmath5 kaon .
we note that the @xmath12 , @xmath194 mass is taken 1400 mev reported by@xcite .
] , @xmath18 , and @xmath193 approximately populate a common trajectory as shown in fig .
5 . the quasi - linear trajectories at the ( n , mass - squared)-plots turned out to be able to describe the light mesons with a good accuracy@xcite .
5 therefore indicates that the @xmath194 , @xmath18 , and @xmath193 could be the good candidates for the @xmath13 , @xmath21 , and @xmath5 kaons , respectively . the @xmath193 is not related any current experimental candidate .
the predicted decay widths of the @xmath193 are listed in table 3 .
the initial state mass is set to 2153 mev .
the total width of the @xmath193 is predicted to be about 197 mev , and the dominant decay modes of the @xmath193 are expected to be @xmath195 , @xmath196 , @xmath197 and @xmath198 .
these results could be of use in searching the candidate for the @xmath5 kaon experimentally .
cc|ccmode & @xmath136 ( mev)&mode & @xmath136 ( mev ) + @xmath199 & 3.9 & @xmath200 & 1.3 + @xmath201 & 1.3 & @xmath202 & 0.08 + @xmath203 & 0.04 & @xmath204 & 12.0 + @xmath205 & 4.5 & @xmath206 & 1.5 + @xmath207 & 0.09 & @xmath208 & 1.6 + @xmath195 & 29.7 & @xmath209 & 34.5 + @xmath210 & 2.5 & @xmath211 & 1.0 + @xmath212 & 6.3 & @xmath213 & 42.8 + @xmath214 & 14.6 & @xmath215 & 26.7 + @xmath216 & 9.9 & @xmath217 & 2.9 + +
with the assignment of the @xmath4 recently observed by the bes collaboration as the @xmath24 member of the @xmath5 meson nonet , the possibility of the @xmath0 , @xmath1 , @xmath2 , and @xmath3 being the @xmath5 @xmath7 states is discussed . with respect to the @xmath0 , its assignment to the @xmath5 isovector state
is not only favored by its mass , but also by its width .
the assignment of the @xmath1 and @xmath3 as the @xmath5 isoscalar states is not favored by their widths .
both the widths of the @xmath2 and @xmath4 can be reasonably reproduced with the mixing angle lying in the range from about @xmath164 to @xmath165 radians .
the assignment of the @xmath0 , @xmath2 , and @xmath4 as the members of the @xmath5 meson nonet not only leads to that the @xmath2-@xmath4 mixing angle is about @xmath218 radians which naturally accounts for the widths of the @xmath2 and @xmath4 , but also gives that the @xmath5 kaon has a mass of about 2153 mev .
the @xmath12 , @xmath194 , @xmath18 , and @xmath193 approximately populate a common ( n , mass - squared)-trajectory . we tend to conclude that the observed pseudoscalar states @xmath0 , @xmath2 , @xmath4 , together with the unobserved @xmath193 appear to be the good candidates for the members of the @xmath5 meson nonet .
the @xmath193 width is predicted to be about 197 mev , and the dominant decay modes of the @xmath193 are expected to be @xmath219 , @xmath209 , @xmath197 and @xmath198 .
these results could be of use in searching the candidate for the @xmath5 kaon experimentally .
this work is supported in part by hancet under contract no .
2006hancet-02 , and the program for youthful teachers in university of henan province .
yao et al . , j. phys .
g * 33 * , 1 ( 2006 ) .
de - min li and bing ma , phys .
d * 77 * , 074004 ( 2008 ) .
m. ablikim et al . , ( bes collaboration ) , phys . rev .
lett . * 95 * , 262001 ( 2005 ) .
m. ablikim et al . , ( bes collaboration ) , phys .
d * 73 * , 112007 ( 2006 ) .
m. ablikim et al . , ( bes collaboration ) , phys .
b * 662 * , 330 ( 2008 ) .
de - min li and bing ma , phys .
d * 77 * , 094021 ( 2008 ) .
n. v. karsnikov and a. a. pivovarov , phys .
b * 112 * , 397 ( 1982 ) ; a. l. kataev , n. v. karsnikov , and a. a. pivovarov , phys .
b * 123 * , 93 ( 1983 ) ; s. g. gorishny , a. l. kataev , and s. a. larin , phys .
b * 135 * , 457 ( 1984 ) ; a. l. kataev , phys .
nucl . * 68 * , 567 ( 2005 ) . s. s. afonin , a. a. andrianov , a. v. andrianov , and d. espriu , jhep , * 0404 * , 039 ( 2004 ) .
v. v. anisovich , l. g. dakhno , m. a. matveev , v. a. nikonov , and a. v. sarantsev , phys . atom
* 70 * , 450 ( 2007 ) .
r. ricken , m. koll , d. merten , b. c. metsch , and h. r. petry , eur .
j. a * 9 * , 221 ( 2000 ) . v. v. khruschev , yad .
fiz . * 46 * , 219 ( 1987 ) ; v. v. khruschev , yad . fiz.*55 * , 773 ( 1992 ) ; s. v. semenov and v. v. khruschev , yad . fiz . * 56 * , 218 ( 1993 ) ; v. i. savrin and s. v. semenov , phys . lett .
b * 374 * , 159 ( 1996 ) ; v. v. khruschev , hep - ph/0504077
. a. v. anisovich , v. v. anisovich , and a. v. sarantsev , phys .
d * 62 * , 051502(r ) ( 2000 ) .
a. le yaouanc , l. oliver , o. pene , and j - c .
raynal , phys .
d * 8 * , 2223 ( 1973 ) ; phys .
d * 9 * , 1415 ( 1974 ) ; phys .
d * 11 * , 1272 ( 1975 ) ; phys .
b * 71 * , 397 ( 1977 ) ; phys . lett .
b * 72*,57 ( 1977 ) .
a. le yaouanc , l. oliver , o. pene , and j - c .
raynal , hadron transitons in the quark model ( gordon and breach science publishers , new york , 1988 ) .
w. roberts and b. silvestr - brac , few - body syst . * 11 * , 171 ( 1992 ) . h. g. blundel , hep - ph/9608473 .
l. micu , nucl .
b * 10 * , 521 ( 1969 ) .
s. capstick and n. isgur , phys .
d * 34 * , 2809 ( 1986 ) ; s. capstick and w. roberts , phys .
d * 49 * 4570 ( 1994 ) .
p. geiger and e. s. swanson , phys .
d * 50 * , 6855 ( 1994 ) .
blundell and s. godfrey , phys .
d * 53*,3700 ( 1996 ) . h. g. blundell , s. godfrey , and b. phelps , phys .
d * 53 * , 3712 ( 1996 ) .
r. kokoski and n. isgur , phys .
d * 35 * , 907 ( 1987 ) .
e. s. ackleh , t. barnes and e. s. swanson , phys .
d * 54 * , 6811 ( 1996 ) .
t. barnes , f. e. close , p. r. page and e. s. swanson , phys .
d * 55 * , 4157 ( 1997 ) .
t. barnes , n. black and p. r. page , phys .
054014 ( 2003 ) . f. e. close and e. s. swanson , phys .
d * 72 * , 094004 ( 2005 ) .
l. burakovsky and p. r. page , phys .
d * 62 * , 014011 ( 2000 ) ; h. q. zhou , r. g. ping , and b. s. zou , phys .
b * 611 * , 123 ( 2005 ) ; t. barnes , s. godfrey and e. s. swanson , phys .
d * 72 * , 054026 ( 2005 ) ; j. lu , w. z. deng , x. l. chen , and s. l. zhu , phys .
d * 73 * , 054012 ( 2006 ) ; b. zhang , x. liu , w. z. deng , and s. l. zhu , eur .
j. c * 50 * , 617 ( 2007 ) ; f. e. close , c. e. thomas , o. lakhina , and e. s. swanson , phys .
b * 647 * , 159 ( 2007 ) ; o. lakhina and e. s. swanson , phys .
b * 650 * , 159 ( 2007 ) ; c. chen , x. l. chen , x. liu , w. z. deng , and s. l. zhu , phys .
d * 75 * , 094017 ( 2007 ) ; g. j. ding and m. l. yan , phys .
b * 657 * , 49 ( 2007 ) .
j. vijande , f. fernandez and a. valcarce , j. phys .
g * 31 * , 481 ( 2005 ) .
s. godfrey and n. isgur , phys .
d * 32 * , 189 ( 1985 ) .
de - min li , bing ma , xue - chao feng , and hong yu , mod .
a * 20 * , 2497 ( 2005 ) .
d. m. li , h. yu and q. x. shen , j. phys .
g * 27 * , 807 ( 2001 ) .
de - min li , ke - wei wei and hong yu , eur .
j. a * 25 * , 263 ( 2005 ) .
s. okubo , prog .
* 27 * , 949 ( 1962 ) .
j. schwinger , phys .
* 12 * , 237 ( 1964 ) .
g. w. brandenberg et al . , ( lass collaboration ) , phys .
* 36 * , 1239 ( 1976 ) . | the strong decays of the @xmath0 , @xmath1 , @xmath2 , @xmath3 , and @xmath4 as the @xmath5 quark - antiquark states are investigated in the framework of the @xmath6 meson decay model .
it is found that the @xmath0 , @xmath2 , and @xmath4 appear to be the convincing @xmath5 @xmath7 states while the assignment of the @xmath1 and @xmath3 as the @xmath5 isoscalar states is not favored by their widths . in the presence of the @xmath0 , @xmath2 , and @xmath4 being the members of the @xmath5 meson nonet , the @xmath5 kaon is phenomenologically determined to has a mass of about 2153 mev .
the width of this unobserved kaon is expected to be about 197 mev in the @xmath6 decay model .
-1 cm mesons , @xmath6 model * pacs numbers:*14.40 .
cs , 12.39.jh |
the swiftly developing field of spin electronics@xcite has provided strong motivation towards the construction of novel magnetic materials . of particular interest
are half - metals , _ i.e. _ compounds for which only the one spin direction presents a gap at the fermi level @xmath1 , while the other has a metallic character .
such behavior is present in various perovskite structures,@xcite in heusler compounds,@xcite and in dilute magnetic semiconductors.@xcite lately it has been possible to grow binary cras in the zinc - blende ( z - b ) structure epitaxially on gaas ; this compound was found to be half metallic , both by experiment and by relevant calculations.@xcite it has also the great advantage of a high curie temperature @xmath2 , around 400 k.@xcite similar is the @xmath2 of ferromagnetic z - b crsb.@xcite moreover , the growth of nanoscale z - b mnas dots on gaas substrates@xcite has been achieved and cras / gaas multilayers have been fabricated.@xcite evidently , the know - how on the construction of these materials , which perhaps can be viewed as the limiting cases of dilute magnetic semiconductors , is increasing , and they will possibly be at the center of interest of spintronics applications soon , since they combine half - metallicity , high curie temperature , and coherent growth on semiconductor ( sc ) substrates .
the experimental work on these systems seems to lie ahead of the theoretical analysis .
pioneering calculations concerning the bulk or the surfaces have been reported,@xcite but there is still extensive work to be done . one matter to be discussed
is the origin of the half - metallicity ; another is the question whether half - metallicity is preserved at the surfaces ; and a third , in view of the future possible growth of similar compounds based on different sc or transition metals , is the theoretical prediction of the cases where half - metallicity can manifest itself , in order to serve as a helpful guideline to where future experiments can focus .
it is the purpose of this paper to address these three points by first - principles calculations .
the outline of the paper has as follows . in section [ sec : calc ]
we briefly describe our method of calculation . in section [ sec : es ] we address the question of the origin of the half - metallic behavior and discuss the electronic structure of the z - b compounds of transiton elements with @xmath0 atoms ; we examine only the ferromagnetic configuration in view of the conclusions of shirai@xcite on vas , cras , and mnas . in section
[ sec : trends ] we present the results of extensive systematic calculations for the z - b compounds of the form xy , with x = v , cr , mn , and y = n , p , as , sb , s , se , te , for various lattice constants in a ferromagnetic configuration , and we examine in which cases half - metallicity is present ; we also present the calculated equilibrium lattice constants to obtain a feeling for the lattice mismatch if one would grow these compounds on semiconductors . in section [ sec : surf ]
we focus on the half - metallicity of transition metal terminated ( 001 ) surfaces .
we conclude with a summary in section [ sec : summary ] .
to perform the calculations , we have used the vosko , wilk and nusair parameterization@xcite for the local density approximation ( lda ) of the exchange - correlation potential to solve the kohn - sham equations within the full - potential screened korringa - kohn - rostoker ( kkr ) green function method,@xcite where the correct shape of the wigner - seitz cells is accounted for.@xcite the scalar relativistic approximation was used , which takes into account the relativistic effects except that of spin - orbit coupling.@xcite to calculate the charge density , we integrate along a contour on the complex energy plane , which extends from below the @xmath3 states of the @xmath0 atom up to the fermi level , using 42 energy points . for the brillouin zone ( bz ) integration we have used a @xmath4-space grid of 30@xmath530@xmath530 in the full bz ( 752 @xmath4-points in the irreducible wedge ) for the bulk calculations and a @xmath6-space grid 30@xmath530 ( 256 @xmath6-points in the irreducible wedge ) in the two - dimensional full bz for the surface calculations .
we have used a cutoff of @xmath7 for the wavefuctions and green functions .
more details on the surface calculations are described in ref . .
the z - b structure can be easily seen to fit on a bcc lattice , by occupying only half of the sites with the two kinds of atoms and leaving the other half as voids .
viewed in this way , the consecutive lattice sites of bcc in the cubic diagonal are occupied by ga , as , void1 , void2 ( in the case of _
e.g. _ gaas ) .
this is useful for the kkr method because all empty space of the quite open z - b structure is taken into account .
we must also say a few words about state - counting with the kkr method .
the charge is found by adding the partial contributions of @xmath8 , @xmath9 , @xmath10 , and @xmath11 electrons ( if a cutoff of @xmath7 is used ) . in this way
one ignores the contribution to the wavefunction of higher angular momenta , which , although very small , is nonzero .
this has a negligible effect for calculations in a metal , where the missing ( or extra ) charge is regained by a very small readjustment of the fermi level @xmath1 .
however , if there is a band gap at @xmath1 , one finds the charge to deviate slightly from the integer value it should have by state - counting . as a consequence for half metals , altough the densitiy of states ( dos ) and the inspection of the bands show clearly that @xmath1 lies within the gap for minority spin , and hence the magnetic moment per unit cell must be integer , its calculated value deviates slightly from the exact integer value . in our work , the presence of half metallicity has always been judged by inspection of the density of states and of the energy bands , and not from the value of the magnetic moment .
the partial values of the charges and moments , calculated by local orbital summation , are accurate to within 1% .
in this section we shall examine the electronic structure of the compounds , and discuss the origin of the half metallicity and the value of the total moment .
although for arbitrary @xmath4 the wavefunctions do not belong exclusively to a single irreducible representation such as @xmath12 or @xmath13 , it is convenient to retain this terminology for the bands formed by originally @xmath12 or @xmath13 orbitals since they are energetically rather separated
. let us focus on a @xmath14 atom . in the z - b structure ,
the tetrahedral environment allows its @xmath12 states ( @xmath15 , @xmath16 , and @xmath17 ) to hybridize with the @xmath9 states of the four first neighbors ( the @xmath0 atoms ) .
note that in the tetrahedral geometry , the @xmath9 orbitals of the four neighboring sites transform as partners of the @xmath12 irreducible representation when analyzed around the central site .
] this creates a large bonding - antibonding splitting , with the low - lying bonding states being more of @xmath9 character around the @xmath0 neighbors ( we call these bands `` @xmath9-bands '' henceforth ) , and the antibonding being rather of @xmath10 character around the @xmath14 atom ( but see also the discussion below concerning refs . and ) .
the gap formed in between is partly filled by the @xmath13 states of the @xmath14 atom ( @xmath18 and @xmath19 ) ; the position of the gap will be different for majority and minority electrons , due to the exchange splitting .
this symmetry - induced @xmath9-@xmath10 hybridization and bonding - antibonding splitting has been found in the past in the case of transiton metal doped zinc - blende semiconductors;@xcite the splitting has also been refered to as `` @xmath9-@xmath10 repulsion '' , concerning the interaction of the impurity @xmath12 localized states with the semiconductor anion @xmath9 band . in the limit of full substitution of the semiconductor cation by the transition metal ,
one arrives to the cases studied here ; in particular mnte has been studied in ref . , and has been found half - metallic if the ferromagnetic phase is assumed .
the situation can be elucidated by examination of the energy bands of cras , presented in fig .
the @xmath8 states of as are very low in energy and omitted in the figure . around @xmath20 , the @xmath9-bands can be seen .
the next bands are the ones formed by the cr @xmath13 orbitals , around @xmath21 for majority and @xmath22 for minority .
they are quite flat , reflecting the fact that their hybridization with the states of as neighbors is weak ( or even zero , at @xmath23 , due to symmetry ) .
these bands can accommodate two electrons per spin . above them ,
the substantially wider antibonding @xmath9-@xmath12 hybrids appear , starting from @xmath24 for majority and from @xmath25 for minority .
their large bandwidth can be attributed to the strong hybridization and to their high energy position .
the bonding - antibonding splitting is also contributing to the fact that they be pushed up beyond the flat @xmath13 states . as a result ,
the three families of bands do not intermix , but are rather energetically separate . in the majority - spin direction
the bonding - antibonding splitting is smaller , because the transition element @xmath14 states are originally lower in energy , closer to the @xmath9 states of the @xmath0 atom . in the minority spin direction
the states are higher due to the exchange splitting and the gap is around @xmath1 .
band counting gives an integer total magnetic moment of 3 @xmath26 . to the clear bonding - antibonding gap
also the z - b geometry contributes , since it has been previously reported@xcite that in the hexagonal nias geometry mnp , mnas and mnsb show no gap . here
we must refer to two recent publications on zinc - blende mnas@xcite , and mnp , mnas , and mnsb,@xcite which have identified the @xmath9 and @xmath12 bands the other way around ; _
i.e. _ they state that the bonding bands are more of @xmath10-character around mn and the antibonding more of @xmath9 character around the @xmath0 atom .
this is correct for the majority bands in those compounds where the majority transition - element @xmath10 states are originally lower than the anion @xmath9 states , _ e.g. _ in the case of the mnas or cras bonding majority bands . for minority - spin
the situation is the other way around , so that the bonding bands are more @xmath9-like centered around the @xmath0 cation , while the antibonding bands are more @xmath10-like centered around the transition element .
this can be seen by the atom - resolved dos in fig .
[ fig2 ] . for the group - vi compounds ,
the @xmath9 states are initially even lower , so that the @xmath9 character of the bonding states increases for both spins .
, @xmath13 , and @xmath12 we denote the whole bands originating from orbitals of the corresponding character , although this character is well defined only for @xmath23 .
the bonding - antibonding splitting between the @xmath9 and @xmath12 bands is evident ; the band gap is between the @xmath9 and the flat @xmath13 bands .
the system is half - metallic . ]
the small dos of the vacant sites is omitted . in all cases
, @xmath1 lies within the minority - spin gap , _
i.e. _ the systems are half - metallic . in the upper left inset
, the majority dos for vas is shown in more detail around @xmath1 , to demonstrate the low dos at @xmath1 .
spin moments for these compounds are given in table .
] next we consider different compounds , to see the trend of increasing transition metal atomic number : vas , cras , mnas .
the corresponding dos can be seen in fig .
[ fig2 ] , calculated for the inas lattice parameter in all cases to allow comparison . for vas , @xmath1 lies between the majority @xmath13 and @xmath12 . since they are separated , this results in zero dos at @xmath1 , as in the limiting case of a semiconductor with almost zero band gap ( shown in more detail in the corresponding inset of fig .
[ fig2 ] ) .
the moment is @xmath27 . as the number of valence @xmath14 electrons increases for cras ,
the @xmath12 bands are pulled down enough to accommodate one electron ( @xmath28 ) , and for mnas even more , so as to accommodate two electrons , whence @xmath29 . at the same time , for the minority spin the band gap does not change in magnitude and continues to embrace @xmath1 .
this can be understood in terms of the exchange splitting @xmath30 being proportional to the magnetic moment : @xmath31 .
when the moment increases by 1 @xmath26 , @xmath30 must increase by one time the exchange integral @xmath32 , about @xmath33.@xcite thus the minority bands remain almost at their position .
this behavior is quite stable , unless the energy gained by increasing the exchange splitting in order to achieve one more @xmath26 is overshot by the coulomb energy cost ; remember that the majority @xmath12 bands are wide .
then the system compromises by placing @xmath1 higher , within the minority @xmath13 conduction band , and half metallicity is lost .
this can occur for small lattice constants or at the surface , and will be presented in the next sections . generally , the bonding @xmath9 bands can accommodate 6 electrons ( 3 per spin ) , and since they lie low in energy they must be filled up ; in addition , there is yet another band lower in energy , consisting of the @xmath8 states of the @xmath0 atom , which has space for two electrons ( one per spin ) .
once these 8 bands are filled by 8 of the valence electrons , the remaining electrons will fill part of majority @xmath10 states , first the lower - lying @xmath13 and then the @xmath12 , creating the magnetic moment .
this electron counting gives then an integer total moment of @xmath34 where @xmath35 is the total number of valence electrons in the unit cell .
, where @xmath36 and @xmath37 are respectively the numbers of majority and minority valence electrons .
] this `` rule of 8 '' is the analogue of the slater - pauling behavior for these systems ; similar behavior is found in the case of the half - metallic heusler alloys , as a `` rule of 18 '' or a `` rule of 24''.@xcite of course these are valid only when half - metallicity is present . according to the `` rule of 8 '' of eq .
( [ eq : sp ] ) , the ( integer ) moment should increase by @xmath38 if one changes only the @xmath0 atom from group v to group vi , since then one more valence electron is available ( in this respect there is a similarity to the dilute magnetic semiconductors@xcite ) . in order to present this , we calculated the electronic structure of the @xmath14group vi compounds . in fig .
[ fig2 ] the dos of crs , crse , and crte are shown .
we chose to use the inas lattice constant in this case , so as to compare with the group v compounds at a single lattice constant .
note that the inas lattice ( 6.058 ) is very close to the cdse one ( 6.052 ) .
we found a ferromagnetic half - metallic solution for these cases , with a magnetic moment of 4 @xmath26 as expected .
the physical mechanism involved is the same as in the @xmath14group v case , and the main difference lies in that @xmath9 states of the group vi elements lie lower in energy , so that the band gap is larger and the hybridization with the neighbor @xmath10 states smaller
. we have also calculated the @xmath14group vi compounds vs , vse , vte , mns , mnse , and mnte .
they also show half - metallic behavior for a range of lattice constants , with their magnetic moment given by the `` rule of 8 '' ; they will be discussed in the next section .
however , we must note that we have only searched for the ferromagnetic solution .
a systematic investigation of the energetics of ferromagnetic _ vs. _
antiferromagnetic solutions lies beyond the scope of this paper . especially in the case of mn
group vi compounds , an antiferromagnetic solution could be possibly more stable , since these compounds are isoelectronic with feas , which has been found to have an antiferromagnetic ground state at the equilibrium lattice constant in calculations by shirai;@xcite especially mnte has been calculated to be more stable in the antifferomagnetic phase , in ref . .
for the compounds vas , cras , and mnas , the analysis of shirai points towards stable ferromagnetic solutions .
.[table:3 ] spin magnetic moment in @xmath26 for the inas experimental lattice constant .
the atomic - resolved moment values refer to the moment included in a wigner - seitz cell around each atom .
void1 and void2 refer to the vacant sites used to describe the zinc - blende structure as described in section .
the total moment does not appear exactly integer as explained in section .
[ cols=">,>,>,>,>,>",options="header " , ] similar trends are observed in the case of @xmath14group vi compounds .
the systematics are presented in table [ table:2 ] .
the investigated systems are xy with x = v , cr , and mn , and y = s , se , and te , for the experimental lattice constants of the ii - vi semiconductors zns , znse , cds , cdse , znte , and cdte .
again we see the loss of half - metallicity by compression , and a stronger tendency towards half - metallicity for lighter group vi elements ; both effects can be attributed to the same reasons as in the @xmath14group v case .
however , the calculated equilibrium lattice constants ( given in parentheses in tables [ table:1 ] and [ table:2 ] ) lessen the possible selections .
indeed , the nitrides and sulphides tend to be so much condensed that they would probably not grow epitaxially at a lattice constant favouring half - metallicity ; the nitrides are even nonmagnetic at their small equilibrium lattice constant .
in fact , only vsb , crsb and crte are half - metallic at their equilibrium lattice constant . on the other hand ,
if one would allow for a 5% expansion due to the lda overbinding or tollerable lattice mismatch , one has the possible combinations vas / gaas , vsb / inas , cras / gaas , crsb / inas , vte / cdte , crse / cds , crte / znte .
other authors have reported results on the equilibrium lattice constant of mnas , cras,@xcite vas,@xcite and crsb.@xcite for mnas , the most frequently studied of these , the reported lda results are : 5.85 (ref . ) , 5.6 - 5.7 (ref . ) , 5.45 (ref . ) , and 5.34 (ref . ) .
our calculations give 5.36 .
we do not know the reason for the discrepancies among the reported values . changing the @xmath39 cutoff or the number of @xmath4-points brought only very small changes to our result ,
while using non - relativistic treatment for the valence electrons resulted in an increase to 5.41 .
results using the generalised gradient approximation ( gga ) as an improvement to the lda have been reported in refs . , , and , giving slightly larger lattice constants compared with lda .
in the last part of this paper we examine the possibility of half - metallicity at the surfaces of these compounds .
one of the authors ( ig ) has recently studied the ( 001 ) surface of cras.@xcite the conclusion of that study was that the as terminated surface loses the half - metallicity due to as dangling bonds that form a surface band within the minority spin gap , while the cr terminated surface retains the half - metallic character and the local magnetic moment increases .
this behavior was found for both the gaas and inas lattice constants .
here we have extended our calculations to include the vas and mnas ( 001 ) surfaces , for the inas lattice constant .
the atoms were kept in their ideal bulk - like positions , _ i.e. _ , we have calculated the unrelaxed and unreconstructed surface ; we must note that large relaxations or reconstruction could alter the half - metallic character of the surface .
since for the as terminated surfaces the surface states originating from dangling bonds can not be avoided , we will discuss the termination with the @xmath14 atom .
although at the surface the @xmath12 and @xmath13 representations are no longer relevant ( the former splits in two , one containing @xmath40 and one @xmath15 , and the latter also in two , one containing @xmath19 and one @xmath18 ) , we ask for understanding that we continue to use the same terminology in order to make the connection to the bulk states . left pannel : spin- and atom - resolved dos of the v - terminated ( 001 ) surface for the v atom at the surface layer and the as atom at the subsurface layer for the inas lattice constant ( left pannel ) .
the surface dos ( full lines ) and the bulk dos ( dashed lines ) are shown .
the as @xmath8 states are located at around 10 ev below the fermi level and are not shown in the figures
. the energy zero refers to @xmath1 ; negative numbers on the dos axis represent the minority spin .
right pannel : same for the mn - terminated mnas ( 001 ) surface . ] the dos for the vas and mnas ( 001 ) surfaces are shown in fig .
[ fig3 ] , for both the atom at the surface layer ( v or mn ) and at the subsurface layer ( as ) . for comparison ,
the corresponding dos in the bulk are also presented .
evidently the v terminated vas ( 001 ) surface remains half - metallic .
however , there is an important difference in the dos as compared to the bulk case , especially at the surface ( v ) layer .
for the majority spin , the peak at around @xmath41 is strongly reduced , and the same happens for the minority spin at around @xmath42 .
this can be explained by the fact that the surface v atom has a reduced coordination number of only two as neighbors .
then two less as @xmath9 orbitals penetrate the v cell , reducing the local weight of the bonding states . at the same time ,
the majority peak just under the fermi level is enhanced , gaining the weight which has been lost by the bonding states from both spin directions .
note in the dos of the v surface atom that there is no more the clear minimum at @xmath1 for the majority states , which was distinguishing @xmath13 from @xmath12 in the bulk , as the reduced bonding - antibonding splitting allows the latter states to come lower in energy . in terms of electron
counting , the v atoms in the bulk give away 3 of the 5 valence electrons to the @xmath9 band and retain the remaining two to build up the magnetic moment . on the other hand , due to the reduced coordination at the surface
, only 1.5 electron per v atom is given away , thus 3.5 remain to fill up majority states and build up the magnetic moment ; due to the gap , the minority states can not take over any of these extra electrons , unless @xmath1 moves higher into the @xmath13-like band ( this is the case for mn to which we shall turn shortly ) .
the situation is given schematically in fig .
[ figschema ] . at the vas surface .
the arrows represent the donation of v electrons to the valence band generated mainly by the low - lying as @xmath9 orbitals . at the v - terminated ( 001 ) surface
, half the neighbors of the v atom are missing , so an excess of 1.5 electrons remains .
they occupy the majority - spin states , giving an extra 1.5 @xmath26 .
the situation is similar in all @xmath14-group v compounds ; for the @xmath14-group vi compounds the @xmath9 band welcomes in total 2 instead of 3 electrons from each @xmath14 atom , thus the excess moment is 1 instead of 1.5 @xmath26.,width=302 ] the gain in magnetic moment at the surface is accompanied by an increase of the exchange splitting , pushing the antibonding and @xmath13-like minority peaks higher in energy ( from 1.5 ev to 2 ev ) ; otherwise the gap could have decreased due to reduced hybridisation .
the increased magnetic moment must reach @xmath43 in the vicinity of the surface ; we find that @xmath44 are located on the surface layer , about @xmath45 in the vacuum , and the rest adds up to the moment of the subsurface layer .
we see that the integer value of the magnetic moment , which is mandatory in bulk half - metallicity , ceases to be a prerequisite in the vicinity of the surface .
the situation described here is completely analogous to the one of the cras ( 001 ) surface.@xcite there , the magnetic moment of the surface layer was reported to reach 4 @xmath26 ; if the magnetic moment of the vacuum and the subsurface layer is added , one finds a total moment of 4.5 @xmath26 in the vicinity of the surface , _ i.e. _ 1.5 @xmath26 more than in the bulk .
however , things are different for mnas ( 001 ) ( with mn termination ) .
the corresponding dos of the surface mn atom and of the subsurface as atom is shown in fig .
[ fig3 ] ( right panel ) .
manifestly , half - metallicity has been lost .
the difference to the vas and cras cases lies in the high magnetic moment of @xmath46 that mnas already shows in the bulk .
if its moment were to increase by the same mechanism , the surface ( plus vacuum and subsurface layers ) would have to accommodate 5.5 majority spin electrons .
this is energetically unfavorable , since the extra half electron can not be accommodated by the @xmath12 band , so that higher - lying bands would have to be populated . instead ,
as the majority states are lowered , the slow gain in magnetic moment is not enough to increase the exchange splitting substantially , and the minority @xmath10 states also reach @xmath1 . as a net result , majority and
minority dos peaks are substantially lower in energy , half - metallicity is lost , and a total magnetic moment of @xmath47 is present within the first surface layer .
spin- and atom - resolved dos of the cr - terminated cras and crse ( 001 ) surfaces for the cr atom at the surface layer and the as and se atoms at the subsurface layer for the inas lattice constant .
the as and se @xmath8 states are located at around 10 ev below the fermi level and are not shown in the figures .
the surface dos are compared to the bulk calculations ( dashed lines ) .
the energy zero refers to @xmath1 ; negative numbers on the dos axis represent the minority spin . ] we have also made calculations for the x terminated xse(001 ) surfaces , where x = v , cr , and mn , using the inas lattice constant .
the v and cr ones are half - metallic while the mn surface is not half - metallic .
the arguments in the above paragraphs still hold , but now each v , mn or cr atom loses se ( group vi ) neighbors instead of as ( group v ) and thus the electron counting changes . in the bulk each x atom gives away only 2 electrons to the @xmath9 band and not 3 as in the xas compounds .
losing two neighbors means that the surface x atom gives away now only 1 electron , so we should have now 4 electrons for v or 5 electrons for cr to build up the spin moment , _ i.e. _ we should have 1 @xmath26 more than in the bulk .
indeed we find the total spin moment to increase to 4 @xmath26 for vse(001 ) and 5 @xmath26 for crse(001 ) .
the mnse compound can not be half - metallic as this would mean a total spin moment of 6 @xmath26 at the surface which is energetically unfavourable for reasons similar to the mnas surface .
a comparison of the cras(001 ) with the crse(001 ) surface dos , both cr terminated and at the inas lattice constant , is shown in fig .
[ fig4 ] , where the bulk dos is also presented .
we see that the differences in the bulk of the two materials ( larger gap , more weight in the majority states under @xmath1 to account for one more electron ) remain their major differences also in the surfaces .
we have studied zinc - blende compounds of the transition metal elements v , cr , mn with the group v @xmath0 elements n , p , as , sb and the group vi s , se , te , in the ferromagnetic configuration .
they all show a tendency towards half - metallic behavior , _
i.e. _ 100% spin polarization at @xmath1 .
this can be traced back to the bonding - antibonding splitting due to hybridization between the transition element @xmath10 ( @xmath12 ) states and the @xmath0 element @xmath9 states , conspiring with the large exchange splitting which pushes up the minority @xmath10 states .
the total moment per unit cell , if the system is half - metallic , is integer and given by the `` rule of 8 '' .
also , the @xmath0 atoms are found to have an antiparallel local moment with respect to the @xmath14 atoms ; the absolute value of the local moments increases for lighter or less ionized @xmath0 elements .
this is traced back to the degree of localization of the @xmath9 wavefunctions around the @xmath0 atoms .
we have discussed the trends with varying lattice constant , in view of the possibility to grow these materials epitaxially on various semiconductors , and calculated the lda equilibrium lattice constants to obtain a feeling of the lattice mismatch with the possible sc substrates .
compression eventually kills half - metallicity , since @xmath1 finally wanders above the minority gap .
although for compounds involving lighter @xmath0 elements half - metallicity is seems more robust , there equilibrium lattice constants are too small .
thus , the best candidates for half - metallicity from this aspect are vas , vsb , cras , crsb , vte , cras , and crte .
finally , we have examined the behavior of the transition element terminated ( 001 ) surfaces . in most cases
half metallicity is maintained , and the magnetic moment increases because of the missing neighbours where charge would be transfered .
exceptions are the cases where the surface magnetic moment should exceed 5 @xmath26 , for which half metallicity is lost . in view of recent experimental success to grow such compounds , and of their relevance to the field of spintronics , we believe that our work will not only add to the understanding of these systems , but will also contribute to the future realization of spintronics devices .
financial support from the rt network of _ computational magnetoelectronics _ ( contract rtn1 - 1999 - 00145 ) of the european commission is greatfully acknowledged .
the authors would like to thank professor p. h. dederichs for helpful discussions and for a critical reading of the manuscript .
motivating discussions with professor h. katayama - yoshida and dr .
k. sato are also gratefully acknowledged . | we report systematic first - principles calculations for ordered zinc - blende compounds of the transition metal elements v , cr , mn with the @xmath0 elements n , p , as , sb , s , se , te , motivated by recent fabrication of zinc - blende cras , crsb , and mnas .
they show ferromagnetic half - metallic behavior for a wide range of lattice constants .
we discuss the origin and trends of half - metallicity , present the calculated equilibrium lattice constants , and examine the half - metallic behavior of their transition element terminated ( 001 ) surfaces . |
stochastic fluctuations are intrinsic to fluid dynamics because fluids are composed of molecules whose positions and velocities are random at thermodynamic scales . because they span the whole range of scales from the microscopic to the macroscopic @xcite
, fluctuations need to be consistently included in all levels of description .
stochastic effects are important for flows in new microfluidic , nanofluidic and microelectromechanical devices @xcite ; novel materials such as nanofluids @xcite ; biological systems such as lipid membranes @xcite , brownian molecular motors @xcite , nanopores @xcite ; as well as processes where the effect of fluctuations is amplified by strong non - equilibrium effects , such as ultra clean combustion , capillary dynamics @xcite , and hydrodynamic instabilities @xcite .
one can capture thermal fluctuations using direct particle level calculations .
but even coarse - grained particle methods @xcite are computationally expensive because the dynamics of individual particles has time scales significantly shorter than hydrodynamic time scales .
alternatively , thermal fluctuations can be included in the navier - stokes equations through stochastic forcing terms , as proposed by landau and lifshitz @xcite and later extended to fluid mixtures @xcite .
the basic idea of _ fluctuating hydrodynamics _ is to add a _
stochastic flux _ corresponding to each dissipative ( irreversible , diffusive ) flux @xcite .
this ensures that the microscopic conservation laws and thermodynamic principles are obeyed while also maintaining fluctuation - dissipation balance .
specifically , the equilibrium thermal fluctuations have the gibbs - boltzmann distribution dictated by statistical mechanics .
fluctuating hydrodynamics is a useful tool in understanding complex fluid flows far from equilibrium @xcite but theoretical calculations are often only feasible after ignoring nonlinearities , inhomogeneities in density , temperature , and transport properties , surface dynamics , gravity , unsteady flow patterns , and other important effects . in the past decade fluctuating hydrodynamics has been applied to study a number of nontrivial practical problems @xcite ; however , the numerical methods used are far from the comparable state - of - the - art for deterministic solvers .
previous computational studies of the effect of thermal fluctuations in fluid mixtures @xcite have been based on the compressible fluid equations and thus require small time steps to resolve fast sound waves ( pressure fluctuations ) .
recently , some of us developed finite - volume methods for the incompressible equations of fluctuating hydrodynamics @xcite , which eliminate the stiffness arising from the separation of scales between the acoustic and vortical modes @xcite .
for inhomogeneous fluids with non - constant density , diffusive mass and heat fluxes create local expansion and contraction of the fluid and the incompressibility constraint should be replaced by a `` quasi - incompressibility '' constraint @xcite .
the resulting _ low - mach number _ equations have been used for some time to model deterministic flows with thermo - chemical effects @xcite , and several conservative finite - volume techniques have been developed for solving equations of this type @xcite . to our knowledge
, thermal fluctuations have not yet been incorporated in low mach number models . in this work
we extend the staggered - grid , finite - volume approach developed in ref .
@xcite to isothermal mixtures of fluids with unequal densities .
the imposition of the quasi - incompressibility constraint poses several nontrivial mathematical and computational challenges . at the mathematical level ,
the traditional low mach number asymptotic expansions @xcite assume spatio - temporal smoothness of the flow and thus do not directly apply in the stochastic context . at the computational level
, enforcing the quasi - incompressibility or equation of state ( eos ) constraint in a conservative and stable manner requires specialized spatio - temporal discretizations . by careful selection of the analytical form of the eos constraint and the spatial discretization of the advective fluxes we are able to maintain strict local conservation and enforce the eos to within numerical tolerances . in the present work ,
we employ an explicit projection - based temporal discretizations because of the substantial complexity of designing and implementing semi - implicit discretizations of the momentum equation for spatially - inhomogeneous fluids @xcite .
thermal fluctuations exhibit unusual features in systems out of thermodynamic equilibrium .
notably , external gradients can lead to _ enhancement _ of thermal fluctuations and to _ long - range _ correlations between fluctuations @xcite .
sharp concentration gradients present during diffusive mixing lead to the development of macroscopic or _ giant fluctuations _
@xcite in concentration , which have been observed using light scattering and shadowgraphy techniques @xcite .
these experimental studies have found good but imperfect agreement between the predictions of a simplified fluctuating hydrodynamic theory and experiments .
computer simulations are , in principle , an ideal tool for studying such complex time - dependent processes in the presence of nontrivial boundary conditions without making the sort of approximations necessary for analytical calculations , such as assuming spatially - constant density and transport coefficients and spatially - uniform gradients . on the other hand , the multiscale ( more precisely , _ many - scale _ ) nature of the equations of fluctuating hydrodynamics
poses many mathematical and computational challenges that are yet to be addressed .
notably , it is necessary to develop temporal integrators that can accurately and robustly handle the large separation of time scales between different physical processes , such as mass and momentum diffusion .
the computational techniques we develop here form the foundation for incorporating additional physics , such as heat transfer and internal energy fluctuations , phase separation and interfacial dynamics , and chemical reactions .
we begin section [ sec : equations ] by formulating the fluctuating low mach number equations for an isothermal binary fluid mixture .
we present both a traditional pressure ( constrained ) formulation and a gauge ( unconstrained ) formulation .
we analyze the spatio - temporal spectrum of the thermal fluctuations in the linearized equations and demonstrate that the low mach equations eliminate the fast ( sonic ) pressure fluctuations but maintain the correct spectrum of the slow ( diffusive ) fluctuations . in section [ sec : temporalintegration ]
we develop projected runge - kutta schemes for solving the spatially - discretized equations , including a midpoint and a trapezoidal second - order predictor - corrector scheme , and a third - order three - stage scheme . in section [ sec : spatialdiscretization ] we describe a spatial discretization of the equations that strictly maintains the equation of state constraint and also obeys a fluctuation - dissipation balance principle @xcite . in section [ sec : giantfluct ] we study the steady - state spectrum of giant concentration fluctuations in the presence of an applied concentration gradient in a mixture of two dissimilar fluids , and test the applicability of common approximations that neglect spatial inhomogeneities .
in section [ sec : mixingmd ] we study the dynamical evolution of giant interface fluctuations during diffusive mixing of two dissimilar fluids , using both hard - disk molecular dynamics and low mach number fluctuating hydrodynamics .
we find excellent agreement between the two , providing a strong support for the usefulness of the fluctuating low mach number equations as a coarse - grained model of complex fluid mixtures . in section
[ sec : conclusions ] we offer some concluding remarks and point out several outstanding challenges for the future .
several technical calculations and procedures are detailed in appendices .
the compressible equations of fluctuating hydrodynamics were proposed some time ago @xcite and have since been studied and applied successfully to a variety of situations @xcite .
the presence of rapid pressure fluctuations due to the propagation of sound waves leads to stiffness that makes it computationally expensive to solve the fully compressible equations numerically , especially for typical liquids .
it is therefore important to develop fluctuating hydrodynamics equations that capture the essential physics in cases where acoustics can be neglected .
it is important to note that the equations of fluctuating hydrodynamics are to be interpreted as a mesoscopic coarse - grained representation of the mass , momentum and energy transport which occurs at microscopic scales through molecular interactions ( collisions ) . as such
, these equations implicitly contain a mesoscopic coarse - graining length and time scale that is larger than molecular scales @xcite .
while a coarse - graining scale does not appear explicitly in the formal stochastic partial differential equations ( spdes ) written in this section ( but note that it can be if desired @xcite ) , it does explicitly enter in the spatio - temporal discretization described in section [ sec : spatialdiscretization ] through the grid spacing ( equivalently , the volume of the grid , or more precisely , the number of molecules per grid cell ) and time step size .
this changes the appropriate interpretation of convergence of numerical methods to a continuum limit in the presence of fluctuations and nonlinearities @xcite . only for the linearized equations of fluctuating hydrodynamics @xcite
can the formal spdes be given a precise continuum meaning @xcite .
developing coarse - grained models that only resolve the relevant spatio - temporal scales is a well - studied but still _ ad hoc _ procedure that requires substantial _ a priori _ physical insight @xcite .
more precise mathematical mode - elimination procedures @xcite are technically involved and often purely formal , especially in the context of spdes @xcite . here
we follow a heuristic approach to constructing fluctuating low mach number equations , starting from the well - known deterministic low mach equations ( which can be obtained via asymptotic analysis @xcite ) and then adding fluctuations in a manner consistent with fluctuation - dissipation balance . alternatively ,
our low mach number equations can be seen as a formal asymptotic limit in which the noise terms are formally treated as smooth forcing terms ; a more rigorous derivation is nontrivial and is deferred for future work .
the starting point of our investigations is the system of isothermal compressible equations of fluctuating hydrodynamics for the density @xmath0 , velocity @xmath1 , and mass concentration @xmath2 for a mixture of two fluids in @xmath3 dimensions . in terms of mass and momentum densities
the equations can be written as conservation laws @xcite , @xmath4+\rho\v g\nonumber \\
v\right)= & \grad\cdot\left[\rho\chi\left(\grad c+k_{p}\grad p\right)+\m{\psi}\right],\label{llns_primitive}\end{aligned}\ ] ] where @xmath5 is the density of the first component , @xmath6 is the density of the second component , @xmath7 is the equation of state for the pressure at the reference temperature @xmath8 , and @xmath9 is the gravitational acceleration .
temperature fluctuations are neglected in this study but can be accounted for using a similar approach .
the shear viscosity @xmath10 , bulk viscosity @xmath11 , mass diffusion coefficient @xmath12 , and baro - diffusion coefficient @xmath13 , in general , depend on the state .
the baro - diffusion coefficient @xmath13 above [ denoted with @xmath14 in ref .
@xcite , see eq .
( a.17 ) in that paper ] is not a transport coefficient but rather determined from thermodynamics @xcite , @xmath15 where @xmath16 is the chemical potential of the mixture at the reference temperature , @xmath17 , and @xmath18 is the isothermal speed of sound .
the capital greek letters denote stochastic momentum and mass fluxes that are formally modeled as @xcite @xmath19 where @xmath20 is boltzmann s constant , and @xmath21 and @xmath22 are standard zero mean , unit variance random gaussian tensor and vector fields with uncorrelated components , @xmath23 and similarly for @xmath24 . at mesoscopic scales , in typical liquids ,
sound waves are much faster than momentum diffusion and can usually be eliminated from the fluid dynamics description .
formally , this corresponds to taking the zero mach number singular limit @xmath25 of the system ( [ llns_primitive ] ) by performing an asymptotic analysis as the mach number @xmath26 , where @xmath27 is a reference flow velocity .
the limiting dynamics can be obtained by performing an asymptotic expansion in the mach number @xcite . in a deterministic setting
this analysis shows that the pressure can be written in the form @xmath28 where @xmath29 .
the low mach number equations can then be obtained by making the anzatz that the thermodynamic behavior of the system is captured by the reference pressure , @xmath30 , and @xmath31 captures the mechanical behavior while not affecting the thermodynamics .
we note that when the system is sufficiently large or the gravitational forcing is sufficiently strong , assuming a spatial constant reference pressure is not valid . in those cases ,
the reference pressure represents a global hydrostatic balance , @xmath32 ( see @xcite for details of the construction of these types of models ) . here , however , we will restrict consideration to cases where gravity causes negligible changes in the thermodynamic state across the domain . in this case
, the reference pressure constrains the system so that the evolution of @xmath33 and @xmath34 remains consistent with the thermodynamic equation of state @xmath35 this constraint means that any change in concentration ( equivalently , @xmath36 ) must be accompanied by a corresponding change in density , as would be observed in a system at thermodynamic equilibrium held at the fixed reference pressure and temperature .
this implies that variations in density are coupled to variations in composition .
note that we do not account for temperature variations in our isothermal model .
the equation for @xmath36 can be written in primitive ( non - conservation ) form as the concentration equation @xmath37 where the non - advective ( diffusive and stochastic ) fluxes are denoted with @xmath38 note that there is no barodiffusion flux because barodiffusion is of thermodynamic origin ( as seen from ( [ eq : k_p ] ) @xcite ) and involves the gradient of the _ thermodynamic _ pressure @xmath39 . by differentiating the eos constraint along a lagrangian trajectory
we obtain @xmath40 where the solutal expansion coefficient @xmath41 is determined by the specific form of the eos .
equation ( [ eq : drho_dt ] ) shows that the eos constraint can be re - written as a constraint on the divergence of velocity , @xmath42 note that the usual incompressibility constraint is obtained when the density is not affected by changes in concentration , @xmath43 .
when @xmath44 changes in composition ( concentration ) due to diffusion cause local expansion and contraction of the fluid and thus a nonzero @xmath45 .
it is important at this point to consider the boundary conditions . for a closed system , such as a periodic domain or a system with rigid boundaries
, we must ensure that the integral of @xmath45 over the domain is zero .
this is consistent with ( [ eq : div_v ] ) if @xmath46 is constant , so that we can rewrite ( [ eq : div_v ] ) in the form @xmath47 . in this case
@xmath30 does not vary in time .
if @xmath46 is not constant , then for a closed system the reference pressure @xmath30 must vary in time to enforce that the total fluid volume remains constant . here
we will assume that @xmath48 , and we will give a specific example of an eos that obeys this condition . the asymptotic low mach analysis of ( [ llns_primitive ] ) is standard and follows the procedure outlined in ref .
@xcite , formally treating the stochastic forcing as smooth .
this analysis leads to the _ isothermal low mach number _ equations for a binary mixture of fluids in conservation form , @xmath49+\rho\v g\equiv\v f(\rho,\v v , c , t)\label{eq : momentum_eq}\\ \partial_{t}\left(\rho_{1}\right)=-\grad\cdot\left(\rho_{1}\v v\right)+ & \grad\cdot\v f\equiv h(\rho,\v v , c , t)\label{eq : rho1_eq}\\ \partial_{t}\left(\rho_{2}\right)=-\grad\cdot\left(\rho_{2}\v v\right ) & -\grad\cdot\v f\label{eq : rho2_eq}\\ \mbox{such that } \grad\cdot\v v= & -\left(\rho^{-1}\beta\right)\,\grad\cdot\v f\equiv s(\rho , c , t).\label{eq : div_v_constraint}\end{aligned}\ ] ] the gradient of the non - thermodynamic component of the pressure @xmath31 ( lagrange multiplier ) appears in the momentum equation as a driving force that ensures the eos constraint ( [ eq : div_v_constraint ] ) is obeyed .
we note that the bulk viscosity term gives a gradient term that can be absorbed in @xmath31 and therefore does not explicitly need to appear in the equations . by adding the two density equations ( [ eq : rho1_eq],[eq : rho2_eq ] ) we get the usual continuity equation for the total density , @xmath50 our conservative numerical scheme is based on eqs .
( [ eq : momentum_eq],[eq : rho1_eq],[eq : div_v_constraint],[eq : rho_eq ] ) . in appendix
[ sec : linearizedanalysis ] , we apply the standard linearized fluctuating hydrodynamics analysis to the low mach number equations .
this gives expressions for the equilibrium and nonequilibrium static and dynamic covariances ( spectra ) of the fluctuations in density and concentration as a function of wavenumber and wavefrequency .
specifically , the dynamic structure factor in the low mach number approximation has the form @xmath51 the linearized analysis shows that the low mach number equations reproduce the slow fluctuations ( small @xmath52 ) in density and concentration ( central rayleigh peak in the dynamic structure factor @xcite ) as in the full compressible equations ( see section [ sub : compressiblespectra ] ) , while eliminating the fast isentropic pressure fluctuations ( side brillouin peaks ) from the dynamics . the fluctuations in velocity , however , are different between the compressible and low mach number equations . in the compressible equations ,
the dynamic structure factor for the longitudinal component of velocity decays to zero as @xmath53 because it has two sound ( brillouin ) peaks centered around @xmath54 , in addition to the central diffusive ( rayleigh ) peak .
the low mach number equations reproduce the central peak ( slow fluctuations ) correctly , replacing the side peaks with a flat spectrum for large @xmath52 , which is unphysical as it formally makes the velocity white in time .
the low mach equations should therefore be used only for time scales larger than the sound propagation time .
the fact that the velocity fluctuations are white in space and in time poses a further challenge in interpreting the nonlinear low mach number equations , and in particular , numerical schemes may not converge to a sensible limit as the time step goes to zero . in practice , just as the spatial discretization of the equations imposes a spatial smoothing or regularization of the fluctuations , the temporal discretization of the equations imposes a temporal smoothing and filters the problematic large frequencies . in the types of problems we study in this work the problem concentration fluctuations can be neglected , @xmath55 , because the concentration fluctuations are dominated by nonequilibrium effects . if @xmath56 the problematic white - in - time longitudinal component of velocity disappears . in general , the eos constraint ( [ eq : general_eos ] )
is a non - linear constraint . in this work
we consider a specific linear eos , @xmath57 where @xmath58 and @xmath59 are the densities of the pure component fluids ( @xmath60 and @xmath61 , respectively ) , giving @xmath62 it is important that for this specific form of the eos @xmath46 is a material constant independent of the concentration .
the density dependence ( [ eq : beta_simple ] ) on concentration arises if one assumes that the two fluids do not change volume upon mixing .
this is a reasonable assumption for liquids that are not too dissimilar at the molecular level .
surprisingly the eos ( [ eq : eos_quasi_incomp ] ) is also valid for a mixture of ideal gases , since @xmath63 where @xmath64 is molecular mass and @xmath65 is the number density .
this is exactly of the form ( [ eq : eos_quasi_incomp ] ) with @xmath66 and @xmath67 .
even if the specific eos ( [ eq : eos_quasi_incomp ] ) is not a very good approximation over the entire range of concentration @xmath68 , ( [ eq : eos_quasi_incomp ] ) may be a very good approximation over the range of concentrations of interest if @xmath58 and @xmath59 are adjusted accordingly . in this case @xmath58 and @xmath59 are not the densities of the pure component fluids but rather fitting parameters that approximate the true eos in the range of concentrations of interest . for small variations in concentration around some reference concentration @xmath69 and density
@xmath70 one can approximate @xmath71 by a constant and determine appropriate values of @xmath58 and @xmath59 from ( [ eq : beta_simple ] ) and the eos ( [ eq : eos_quasi_incomp ] ) evaluated at the reference state .
our specific form choice of the eos will aid significantly in the construction of simple conservative spatial discretizations that strictly maintain the eos without requiring complicated nonlinear iterative corrections .
several different types of boundary conditions can be imposed for the low mach number equations , just as for the more familiar incompressible equations .
the simplest case is when periodic boundary conditions are used for all of the variables .
we briefly describe the different types of conditions that can be imposed at a physical boundary with normal direction @xmath72 . for the concentration ( equivalently , @xmath36 ) , either neumann ( zero mass flux ) or dirichlet ( fixed concentration ) boundary conditions can be imposed .
physically , a neumann condition corresponds to a physical boundary that is impermeable to mass , while dirichlet conditions correspond to a permeable membrane that connects the system to a large reservoir held at a specified concentration . in the case of neumann conditions for concentration , both the normal component of the diffusive flux @xmath73 and the advective flux @xmath74 vanish at the boundary , implying that the normal component of velocity must vanish , @xmath75 . for dirichlet conditions on the concentration , however , there will , in general , be a nonzero normal diffusive flux @xmath76 through the boundary .
this diffusive flux for concentration will induce a corresponding mass flux , as required to maintain the equation of state near the boundary . from the condition ( [ eq : div_v_constraint ] ) , we infer the proper boundary condition for the normal component of velocity to be @xmath77 this condition expresses the notion that there is no net volume change for the fluid in the domain . note that no additional boundary conditions can be specified for @xmath33 since its boundary conditions follow from those on @xmath34 via the eos constraint . for the tangential component of velocity @xmath78
, we either impose a no - slip condition @xmath79 , or a free slip boundary condition in which the tangential component of the normal viscous stress vanishes , @xmath80 in the case of zero normal mass flux , @xmath75 , the free slip condition simplifies to a neumann condition for the tangential velocity , @xmath81 .
the low mach number system of equations ( [ eq : momentum_eq],[eq : rho1_eq],[eq : div_v_constraint],[eq : rho_eq ] ) is a _ constrained _ problem . for the purposes of analysis and in particular for constructing higher - order temporal integrators , it is useful to rewrite the constrained low mach number equations as an _ unconstrained _ initial value problem . in the incompressible case , @xmath82 , we can write the constrained navier - stokes equations as an unconstrained system by eliminating the pressure using a projection operator formalism . the constraint @xmath82 is a constant linear constraint and independent of the state and of time .
however , in the low mach number equations the velocity - divergence constraint @xmath83 depends on concentration , and also on time when there are additional ( stochastic or deterministic ) forcing terms in the concentration equation . treating this type of system requires a more general vector field decomposition .
this more general vector field decomposition provides the basis for a projection - based discretization of the constrained system .
we also introduce a gauge formulation of the system @xcite that casts the evolution as a nonlocal unconstrained system that is analytically equivalent to the orignal constrained evolution .
the gauge formulation allows us to develop higher - order method - of - lines temporal integration algorithms .
the velocity in the low mach number equations can be split into two components , @xmath84 where @xmath85 is a divergence - free ( solenoidal or vortical ) component , and therefore @xmath86 this is a poisson problem for @xmath87 that is well - posed for appropriate boundary conditions on @xmath88 .
specifically , periodic boundary conditions on @xmath88 imply periodic boundary conditions for @xmath89 and @xmath87 . at physical boundaries where a dirichlet condition ( [ eq : v_n_bc ] ) is specified for the normal component of the velocity , we set @xmath90 and use neumann conditions for the poisson solve , @xmath91
. we can now define a more general vector field decomposition that plays the role of the hodge decomposition in incompressible flow . given a vector field @xmath92 and a density @xmath33 we can decompose @xmath92 into three components
@xmath93 this decomposition can be obtained by using the condition @xmath85 and @xmath94 , which allows us to define a density - weighted poisson equation for @xmath95 , @xmath96 let @xmath97 denote the solution operator to the density - dependent poisson problem , formally , @xmath98^{-1},\ ] ] and also define a density - dependent projection operator @xmath99 defined through its action on a vector field @xmath100 , @xmath101.\ ] ] this is a well - known variable density generalization @xcite of the constant - density projection operator @xmath102 $ ] .
we can now write @xmath103-\grad\zeta.\ ] ] this gives @xmath104 where we have introduced an affine transformation @xmath105 that depends on @xmath33 , @xmath34 and @xmath106 through @xmath107 , and is defined via its action on a vector field @xmath100 , @xmath108.\label{p_tilde_v}\ ] ] note that application of @xmath109 requires only one poisson solve and does not actually require computing @xmath87 .
the low mach number system ( [ eq : momentum_eq],[eq : rho1_eq],[eq : rho_eq],[eq : div_v_constraint ] ) has the form @xmath110 where @xmath111 is the momentum density , and @xmath112 , @xmath113 and @xmath114 are as defined in ( [ eq : momentum_eq],[eq : rho1_eq],[eq : div_v_constraint ] ) . at present
, we will assume that these functions are smooth functions of time , which is only justified in the presence of stochastic forcing terms in a linearized setting .
we note that , for the constrained system , @xmath33 is not an independent variable because of the eos constraint ( [ eq : eos_quasi_incomp ] ) ; however , we will retain the evolution of @xmath33 with the implicit understanding that the evolution must be constrained so that @xmath33 and @xmath34 remain consistent with ( [ eq : eos_quasi_incomp ] ) . to define the gauge formulation , we introduce a new variable @xmath115 where @xmath95 is a _ gauge _ variable .
we note that @xmath95 is not uniquely determined ; however , the specific choice does not matter . if we choose the gauge so that @xmath116 then the momentum equation in ( [ eq : div_v_s ] ) is equivalent to @xmath117 the appropriate boundary conditions for @xmath95 are linked to the boundary conditions on @xmath88 ; we set @xmath95 to be periodic if @xmath88 is periodic , and employ a homogeneous neumann ( natural ) boundary condition @xmath118 if a dirichlet condition ( [ eq : v_n_bc ] ) is specified for the normal component of the velocity @xmath119 .
note that in the spatially - discrete staggered formulation that we employ , the homogeneous neumann condition follows automatically from the boundary conditions on velocity used to define the appropriate divergence and gradient operators in the interior of the domain .
if we know @xmath120 and @xmath33 , we can then define @xmath121 and compute @xmath122 , where @xmath109 is defined in ( [ p_tilde_v ] ) .
thus by using the gauge formulation we can formally write the low mach number equations in the form of an unconstrained initial value problem @xmath123 the utility of the gauge formulation is that in fact , we do not need to know @xmath95 in order to determine @xmath88 .
therefore , the time evolution equation for @xmath95 does not actually need to be solved , and in particular , @xmath31 does not need to be computed .
futhermore , by adopting the gauge formulation , we can directly use a method of lines approach for spatially - discretizing the system ( [ eq : m_gauge],[eq : rho1_gauge ] ) , and then apply standard runge - kutta temporal integrators to the resulting system of ordinary ( stochastic ) differential equations .
it is important to emphasize that the actual independent physical variables in the low mach formulation ( [ eq : m_gauge],[eq : rho1_gauge ] ) are the vortical ( solenoidal ) component of velocity @xmath89 and the concentration @xmath34 .
the density @xmath124 and the velocity @xmath125 $ ] are determined from @xmath89 and @xmath34 and the constraints ; hence they can formally be eliminated from the system , as can be seen in the linearized analysis in appendix [ sec : linearizedanalysis ] , which shows that fluctuations in the vortical velocity modes are decoupled from the longitudinal fluctuations .
our spatio - temporal discretization follows a `` method of lines '' approach in which we first discretize the equations ( [ eq : momentum_eq],[eq : rho1_eq],[eq : div_v_constraint],[eq : rho_eq ] ) in space and then integrate the resulting semi - continuum equations in time .
our uniform staggered - grid spatial discretization of the low mach number equations is relatively standard and is described in section [ sec : spatialdiscretization ] .
the main difficulty is the temporal integration of the resulting equations in the presence of the eos constraint .
our temporal integrators are based on the gauge formulation ( [ eq : m_gauge],[eq : rho1_gauge ] ) of the low mach equations .
the gauge formulation is unconstrained and enables us to use standard temporal integrators for initial - value problems . in the majority of this section ,
we assume that all of the fields and differential operators have already been spatially discretized and focus on the temporal integration of the resulting initial - value problem .
because in the present schemes we handle both diffusive and advective fluxes explicitly , the time step size @xmath126 is restricted by well - known cfl conditions . for fluctuating hydrodynamics applications the time step is typically limited by momentum diffusion , @xmath127 where @xmath3 is the number of spatial dimensions and @xmath128 is the grid spacing .
the design and implementation of numerical methods that handle momentum diffusion semi - implicitly , as done in ref .
@xcite for incompressible flow , is substantially more difficult for the low mach number equations because it requires a variable coefficient implicit fluid solver .
we have recently developed an efficient stokes solver for solving variable - density and variable - viscosity time - dependent and steady stokes problems @xcite , and in future work we will employ this solver to construct a semi - implicit temporal integrator for the low mach number equations . our temporal discretization will make use of the special form of the eos and the discretization of mass advection described in section [ sub : advection ] in order to strictly maintain the eos relation ( [ eq : eos_quasi_incomp ] ) between density and concentration in each cell at _ all _ intermediate values .
therefore , no additional action is needed to enforce the eos constraint after an update of @xmath36 and @xmath33 .
this is , however , only true to within the accuracy of the poisson solver and also roundoff , and it is possible for a slow drifting off the eos to occur over many time steps . in section [ sub : driftcorrection ] , we describe a correction that prevents such drifting and ensures that the eos is obeyed at all times to essentially roundoff tolerance . for simplicity , we will often omit the explicit update for the density @xmath33 and instead focus on updating @xmath36 and the momentum density @xmath111 , with the understanding that @xmath33 is updated whenever @xmath36 is .
the foundation for our higher - order explicit temporal integrators is the first - order euler method applied to the gauge formulation ( [ eq : m_gauge],[eq : rho1_gauge ] ) .
we use a superscript to denote the time step and the point in time where a given term is evaluated , e.g. , @xmath129 where @xmath130 denotes the spatial discretization of @xmath131 with analogous definitions for @xmath132 and @xmath133 .
we also denote the time step size with @xmath134 .
assume that at the beginning of timestep @xmath72 we know @xmath135 and we can then compute @xmath136 by enforcing the constraint ( [ eq : div_v_s ] ) .
here @xmath137 denotes the affine transformation ( [ p_tilde_v ] ) with all terms evaluated at the beginning of the time step , so that @xmath138 . an euler step for the low mach equations
then consists of the update @xmath139 together with an update of the density @xmath140 consistent with @xmath141 . at the beginning of the next time step ,
@xmath142 will be calculated from @xmath143 by applying @xmath144 , and it is only @xmath142 that will actually be used during time step @xmath145 .
we therefore do not need to explicitly store @xmath143 and can instead replace it with @xmath146 without changing any of the observable results .
this is related to the fact that the gauge is _ de facto _ arbitrary and , in the present setting , the gauge formulation is simply a formalism to put the equations in an unconstrained form suitable for method of lines discretization .
the difference between @xmath120 and @xmath147 is a ( discrete ) gradient of a scalar .
since our temporal integrators only use linear combinations of the intermediate values , the difference between the final result for @xmath143 and @xmath148 is also a gradient of a scalar and replacing @xmath143 with @xmath149 simply amounts to redefining the ( arbitrary ) gauge variable . for these reasons
, the euler advance , @xmath150,\label{eq : euler_step_complex}\end{aligned}\ ] ] is analytically equivalent to ( [ eq : euler_lagged ] ) .
we will use this form as the foundation for our temporal integrators .
the equivalence to the gauge form implies that the update specified by ( [ eq : euler_step_complex ] ) can be viewed as an explicit update in spite of the formal dependence of the update on the solution at both old and new time levels .
thermal fluctuations can not be straightforwardly incorporated in ( [ eq : euler_step_complex ] ) because it is not clear how to define @xmath144 . in the deterministic setting ,
@xmath114 is a function of concentration and density and can be evaluated pointwise at time level @xmath145 .
when the white - in - time stochastic concentration flux @xmath151 is included , however , @xmath114 can not be evaluated at a particular point of time .
instead , one must think of @xmath151 as representing the _ average _ stochastic flux over a given time interval @xmath152 , which can be expressed in terms of the increments @xmath153 of the underlying wiener processes , @xmath154 where @xmath155 is a collection of normal variates generated using a pseudo - random number generator , and @xmath156 is the volume of the hydrodynamic cells .
similarly , the average stochastic momentum flux over a time step is modeled as @xmath157 where @xmath158 are normal random variates . as described in more detail in ref .
@xcite , stochastic fluxes are spatially discretized by generating normal variates on the faces of the grid on which the corresponding variable is discretized , independently at each time step . as mentioned earlier , the volume of the grid cell appears here because it expresses the spatial coarse graining length scale ( i.e. , the degree of coarse - graining for which a fluid element with discrete molecules can be modeled by continuous density fields ) implicit in the equations of fluctuating hydrodynamics .
similarly , the time interval @xmath159 expresses the typical time scale at which the mass and momentum transfer can be modeled with low mach number hydrodynamics . with this in mind
, we first evaluate the velocity divergence associated with the constraint using the particular sample of @xmath151 , @xmath160.\ ] ] we then define a discrete affine operator @xmath161 in terms of its action on the momentum @xmath162
@xmath163\left(\v m\right)=\rho\m{\mathcal{r}}_{s}\left(\rho^{-1}\v m\right).\ ] ] using this shorthand notation , the momentum update in ( [ eq : euler_step_complex ] ) in the presence of thermal fluctuations can be written as @xmath164\left(\v m^{n}+\d t\,\v f^{n}\right).\ ] ] observe that this is a conservative momentum update since the application of @xmath165 subtracts the ( discrete ) gradient of a scalar from the momentum . in actual implementation , it is preferable to apply @xmath166 at the beginning of the time step @xmath145 instead of at the end of time step @xmath72 , once the value @xmath167 is computed from the diffusive and stochastic fluxes for the concentration . following the above discussion , we can write an euler - maruyama temporal integrator for the low mach number equations in the shorthand notation , @xmath168\left(\tilde{\v m}^{n}\right)\nonumber \\ \rho_{1}^{n+1 } & = & \rho_{1}^{n}+\d t\,\bar{h}^{n}+\check{h}^{n}\left(\d t,\,\widetilde{\v w}^{n}\right)\nonumber \\ \tilde{\v m}^{n+1 } & = & \v m^{n}+\d t\,\bar{\v f}^{n}+\check{\v f}^{n}\left(\d t,\,\m w^{n}\right),\label{eq : euler_step}\end{aligned}\ ] ] where @xmath169 and @xmath170 are collections of standard normal variates generated using a pseudo - random number generator independently at each time step . here
the deterministic increments are written using the shorthand notation , @xmath171+\rho\v g\\ \bar{h } & = & \grad\cdot\left(-\rho_{1}\v v+\rho\chi\grad c\right).\end{aligned}\ ] ] the stochastic increments are written in terms of @xmath172\delta t=\grad\cdot\left[\sqrt{\frac{\eta\left(k_{b}t\right)\d t}{\d v}}\,\left(\m w+\m w^{t}\right)\right]\\ \check{h}\left(\d t,\,\widetilde{\v w}\right ) & = & \left[\grad\cdot\m{\psi}\left(\d t,\,\widetilde{\v w}\right)\right]\delta t=\grad\cdot\left[\sqrt{\frac{2\chi\rho\mu_{c}^{-1}\left(k_{b}t\right)\d t}{\d v}}\,\widetilde{\v w}\right],\end{aligned}\ ] ] where @xmath155 and @xmath158 are vectors of standard gaussian variables @xcite .
a good strategy for composing higher - order temporal integrators for the low mach number equations is to use a linear combination of several projected euler steps of the form ( [ eq : euler_step ] ) . in this way
, the higher - order integrators inherit the properties of the euler step . in our case
, this will be very useful in constructing conservative discretizations that strictly maintain the eos constraint and only evaluate fluxes at states that strictly obey the eos constraint .
the incorporation of stochastic forcing in the runge - kutta temporal integrators that we use is described in refs .
@xcite ; here we only summarize the resulting schemes .
we note that the stochastic terms should be considered additive noise , even though we evaluate them using an instantaneous state like multiplicative noise @xcite .
a weakly second - order temporal integrator for ( [ eq : m_gauge],[eq : rho1_gauge ] ) is provided by the _ explicit trapezoidal rule _
, in which we first take a predictor euler step @xmath168\left(\tilde{\v m}^{n}\right)\nonumber \\
\rho_{1}^{\star , n+1 } & = & \rho_{1}^{n}+\d t\,\bar{h}^{n}+\check{h}^{n}\left(\d t,\,\widetilde{\v w}^{n}\right)\\ \tilde{\v m}^{\star , n+1 } & = & \v m^{n}+\d t\,\bar{\v f}^{n}+\check{\v f}^{n}\left(\d t,\,\m w^{n}\right).\label{eq : trapezoidal_predictor}\end{aligned}\ ] ] the corrector step is a linear combination of the predictor and another euler update , @xmath173\left(\tilde{\v m}^{\star , n+1}\right)\nonumber \\ \rho_{1}^{n+1 } & = & \frac{1}{2}\rho_{1}^{n}+\frac{1}{2}\left[\rho_{1}^{\star , n+1}+\d t\,\bar{h}^{\star , n+1}+\check{h}^{\star , n+1}\left(\d t,\,\widetilde{\v w}^{n}\right)\right]\\ \tilde{\v m}^{n+1 } & = & \frac{1}{2}\v m^{n}+\frac{1}{2}\left[\m m^{\star ,
n+1}+\d t\,\bar{\v f}^{\star , n+1}+\check{\v f}^{\star , n+1}\left(\d t,\,\m w^{n}\right)\right],\label{eq : trapezoidal_corrector}\end{aligned}\ ] ] and reuses the same random numbers @xmath169 and @xmath170 as the predictor step . note that both the predicted and the corrected values for density and concentration obey the eos .
we numerically observe that the trapezoidal rule does exhibit a slow but systematic numerical drift in the eos , and therefore it is necessary to use the correction procedure described in section [ sub : driftcorrection ] at the end of each time step .
the analysis in ref .
@xcite indicates that for the incompressible case the trapezoidal scheme exhibits second - order weak accuracy in the nonlinear and linearized settings .
an alternative second - order scheme is the _ explicit midpoint rule _ , which can be summarized as follows .
first we take a projected euler step to estimate midpoint values ( denoted here with superscript * @xmath174 * ) , @xmath175\left(\tilde{\v m}^{n}\right)\nonumber \\ \rho_{1}^{\star , n+\myhalf } & = & \rho_{1}^{n}+\frac{\d t}{2}\,\bar{h}^{n}+\check{h}^{n}\left(\frac{\d t}{2},\,\widetilde{\v w}_{1}^{n}\right)\nonumber \\
\tilde{\m m}^{\star , n+\myhalf } & = & \v m^{n}+\frac{\d t}{2}\,\bar{\v f}^{n}+\check{\v f}^{n}\left(\frac{\d t}{2},\,\m w_{1}^{n}\right).\label{eq : midpoint_predictor}\end{aligned}\ ] ] and then we complete the time step with another euler - like update @xmath176\left(\tilde{\v m}^{\star , n+\myhalf}\right)\nonumber \\ \rho_{1}^{n+1 } & = & \rho_{1}^{n}+\d t\,\bar{h}^{\star , n+\myhalf}+\check{h}^{\star , n+\myhalf}\left(\d t,\,\widetilde{\v w}^{n}\right)\nonumber \\
\tilde{\v m}^{n+1 } & = & \v m^{n}+\d t\,\bar{\v f}^{\star , n+\myhalf}+\check{\v f}^{\star , n+\myhalf}\left(\d t,\,\m w^{n}\right),\label{eq : midpoint_corrector}\end{aligned}\ ] ] where the standard gaussian variates @xmath177 and the vectors of standard normal variates @xmath178 and @xmath179 are independent , and similarly for @xmath180 and @xmath181 .
note that @xmath178 and @xmath180 are used in _ both _ the predictor and the corrector stages , while @xmath179 and @xmath181 are used in the corrector only .
physically , the random numbers @xmath182 ( and similarly for @xmath178 ) correspond to the increments of the underlying wiener processes @xmath183 over the first half of the time step , and the random numbers @xmath184 correspond to the wiener increments for the second half of the timestep @xcite .
note that both the midpoint and the endpoint values for density and concentration obey the eos .
we numerically observe that the midpoint rule does not exhibit a systematic numerical drift in the eos , and can therefore be used without the correction procedure described in section [ sub : driftcorrection ] . the analysis in ref .
@xcite indicates that for the incompressible case the midpoint scheme exhibits second - order weak accuracy in the nonlinear setting .
furthermore , in the linearized setting it reproduces the steady - state covariances of the fluctuating fields to third order in the time step size .
we have also tested and implemented the three - stage runge kutta scheme that was used in refs .
this scheme can be expressed as a linear combination of three euler steps .
the first stage is a predictor euler step , @xmath168\left(\tilde{\v m}^{n}\right)\nonumber \\
\rho_{1}^{\star } & = & \rho_{1}^{n}+\d t\,\bar{h}^{n}+\check{h}^{n}\left(\d t,\,\widetilde{\v w}^{n}\right)\\ \tilde{\v m}^{\star } & = & \v m^{n}+\d t\,\bar{\v f}^{n}+\check{\v f}^{n}\left(\d t,\,\m w^{n}\right).\label{eq : trapezoidal_predictor-1}\end{aligned}\ ] ] the second stage is a midpoint predictor @xmath185\left(\tilde{\v m}^{\star}\right)\nonumber \\
\rho_{1}^{\star\star } & = & \frac{3}{4}\rho_{1}^{n}+\frac{1}{4}\left[\rho_{1}^{\star}+\d t\,\bar{h}^{\star}+\check{h}^{\star}\left(\d t,\,\widetilde{\v w}^{\star , n}\right)\right]\\ \tilde{\v m}^{\star\star } & = & \frac{3}{4}\v m^{n}+\frac{1}{4}\left[\m m^{\star}+\d t\,\bar{\v f}^{\star}+\check{\v f}^{\star}\left(\d t,\,\m w^{\star , n}\right)\right],\label{eq : rk3_predictor}\end{aligned}\ ] ] and a final corrector stage completes the time step @xmath186\left(\tilde{\v m}^{\star\star}\right)\nonumber \\ \rho_{1}^{n+1 } & = & \frac{1}{3}\rho_{1}^{n}+\frac{2}{3}\left[\rho_{1}^{\star\star}+\d t\,\bar{h}^{\star\star}+\check{h}^{\star\star}\left(\d t,\,\widetilde{\v w}^{\star\star , n}\right)\right]\\ \tilde{\m m}^{n+1 } & = & \frac{1}{3}\v m^{n}+\frac{2}{3}\left[\m m^{\star\star}+\d t\,\bar{\v f}^{\star\star}+\check{\v f}^{\star\star}\left(\d t,\,\m w^{\star\star , n}\right)\right].\label{eq : rk3_corrector}\end{aligned}\ ] ] here the stochastic fluxes between different stages are related to each other via @xmath187 where @xmath188 and @xmath189 are independent and generated independently at each rk3 step , and similarly for @xmath155 .
the weights of @xmath189 are chosen to maximize the weak order of accuracy of the scheme while still using only two random samples of the stochastic fluxes per time step @xcite .
the rk3 method is third - order accurate deterministically , and stable even in the absence of diffusion / viscosity ( i.e. , for advection - dominated flows ) .
note that the predicted , the midpoint and the endpoint values for density and concentration all obey the eos .
we numerically observe that the rk3 scheme does exhibit a systematic numerical drift in the eos , and therefore it is necessary to use the correction procedure described in section [ sub : driftcorrection ] at the end of each time step .
the analysis in ref .
@xcite indicates that for the incompressible case the rk3 scheme exhibits second - order weak accuracy in the nonlinear setting . in the linearized
setting it reproduces the steady - state covariances of the fluctuating fields to third order in the time step size . while in principle our temporal integrators
should strictly maintain the eos , roundoff errors and the finite tolerance employed in the iterative poisson solver lead to a small drift in the constraint that can , depending on the specific scheme , lead to an exponentially increasing violation of the eos over many time steps . in order to maintain the eos at all times to within roundoff tolerance
, we periodically apply a globally - conservative @xmath190 projection of @xmath33 and @xmath36 onto the linear eos constraint .
this projection step consists of correcting @xmath36 in cell @xmath191 using @xmath192+\frac{1}{n}\sum_{k^{\prime}}\left(\rho_{1}\right)_{k^{\prime}},\ ] ] where @xmath193 is the number of hydrodynamic cells in the system and @xmath194 note that the above update , while nonlocal in nature , conserves the total mass @xmath195 .
a similar update applies to @xmath196 , or equivalently , @xmath197 .
the spatial discretization we employ follows closely the spatial discretization of the constant - coefficient incompressible equations described in ref .
therefore , we focus here on the differences , specifically , the use of conserved variables , the handling of the variable - density projection and variable - coefficient diffusion , and the imposition of the low mach number constraint . note that the handling of the stochastic momentum and mass fluxes is identical to that described in ref .
@xcite . for simplicity of notation , we focus on two dimensional problems , with straightforward generalization to three spatial dimensions .
our spatial discretization follows the commonly - used mac approach @xcite , in which the scalar conserved quantities @xmath33 and @xmath36 are defined on a regular cartesian grid .
the vector conserved variables @xmath111 are defined on a staggered grid , such that the @xmath198 component of momentum is defined on the faces of the scalar variable cartesian grid in the @xmath198 direction , see fig .
[ fig : grid ] . for simplicity of notation
, we often denote the different components of velocity as @xmath199 in two dimensions and @xmath200 in three dimensions .
the terms `` cell - centered '' , `` edge - centered '' , and `` face - centered '' refer to spatial locations relative to the underlying scalar grid .
our discretization is based on calculating fluxes on the faces of a finite - volume grid and is thus locally conservative .
it is important to note , however , that for the mac grid different control volumes are used for the scalars and the components of the momentum , see fig .
[ fig : grid ] . staggered ( mac ) finite - volume discretization on a uniform cartesian two - dimensional grid .
( _ left _ ) control volume and flux discretization for cell - centered scalar fields , such as densities @xmath33 and @xmath36 .
( _ middle _ ) control volume for the @xmath201-component of face - centered vector fields , such as @xmath202 ( _ right _ ) control volume for the @xmath203-component of face - centered vector fields , such as @xmath204 .
, width=576 ] from the cell - centered @xmath33 and @xmath36 we can define other cell - centered scalar quantities , notably , the concentration @xmath205 and the transport quantities @xmath206 and @xmath207 , which typically depend on the local density @xmath208 and concentration @xmath209 ( and temperature for non - isothermal models ) , and can , in general , also depend on the spatial position of the cell @xmath210 . in order to define velocities
we need to interpret the continuum relationship @xmath111 on the staggered grid .
this is done by defining face - centered scalar quantities obtained as an arithmetic average of the corresponding cell - centered quantities in the two neighboring cells .
specifically , we define @xmath211 except at physical boundaries , where the value is obtained from the imposed boundary conditions ( see section [ sub : boundaryconditions ] ) .
arithmetic averaging is only one possible interpolation from cells to faces @xcite . in general , other forms of averaging such as a harmonic or geometric average or higher - order ,
wider stencils @xcite can be used .
most components of the spatial discretization can easily be generalized to other choices of interpolation . as we explain later
, the use of linear averaging simplifies the construction of conservative advection . in this section
we describe the spatial discretization of the diffusive mass flux term @xmath212 in ( [ eq : rho1_eq ] ) .
the discretization is based on conservative centered differencing @xcite , @xmath213+\delta y^{-1}\left[\left(\rho\chi\frac{\partial c}{\partial y}\right)_{i , j+\myhalf}-\left(\rho\chi\frac{\partial c}{\partial y}\right)_{i , j-\myhalf}\right],\label{eq : discrete_diffusion}\ ] ] where , for example , @xmath214 and @xmath215 is an interpolated face - centered diffusion coefficient , for example , as done for @xmath33 in eq .
( [ eq : rho_face ] ) , @xmath216 except at physical boundaries , where the value is obtained from the imposed boundary conditions .
regardless of the specific form of the interpolation operator , the same face - centered diffusion coefficient @xmath215 must be used when calculating the magnitude of the stochastic mass flux on face @xmath217 , @xmath218 this matches the covariance of the discrete stochastic mass increments @xmath219 with the discretization of the diffusive dissipation operator @xmath220 given in ( [ eq : discrete_diffusion],[eq : discrete_diff_flux ] ) .
this matching ensures discrete fluctuation - dissipation balance in the linearized setting @xcite .
specifically , at thermodynamic equilibrium the static covariance of the concentration is determined from the equilibrium value of @xmath221 ( thermodynamics ) independently of the particular values of the transport coefficients ( dynamics ) , as seen in ( [ eq : s_equilibrium ] ) and dictated by statistical mechanics principles . in ref .
@xcite a laplacian form of the viscous term @xmath222 is assumed , which is not applicable when viscosity is spatially varying and @xmath223 . in two dimensions ,
the divergence of the viscous stress tensor in the momentum equation ( [ eq : momentum_eq ] ) , neglecting bulk viscosity effects , is @xmath224 & = & \left[\begin{array}{c } 2\frac{\partial}{\partial x}\left(\eta\frac{\partial u}{\partial x}\right)+\frac{\partial}{\partial y}\left(\eta\frac{\partial u}{\partial y}+\eta\frac{\partial v}{\partial x}\right)\\ 2\frac{\partial}{\partial y}\left(\eta\frac{\partial v}{\partial y}\right)+\frac{\partial}{\partial x}\left(\eta\frac{\partial v}{\partial x}+\eta\frac{\partial u}{\partial y}\right ) \end{array}\right].\label{eq : div_viscous_stress}\end{aligned}\ ] ] the discretization of the viscous terms requires @xmath10 at cell - centers and edges ( note that in two dimensions the edges are the same as the nodes @xmath225 of the grid ) .
the value of @xmath10 at a node is interpolated as the arithmetic average of the four neighboring cell - centers , @xmath226 except at physical boundaries , where the values are obtained from the prescribed boundary conditions .
the different viscous friction terms are discretized by straightforward centered differences .
explicitly , for the @xmath201-component of momentum @xmath227_{i+\myhalf , j}=\d x^{-1}\left[\left(\eta\frac{\partial u}{\partial x}\right)_{i+1,j}-\left(\eta\frac{\partial u}{\partial x}\right)_{i , j}\right]\ ] ] with @xmath228 similarly , for the term involving a second derivative in @xmath203 , @xmath229_{i+\myhalf , j}=\d y^{-1}\left[\left(\eta\frac{\partial u}{\partial y}\right)_{i+\myhalf , j+\myhalf}-\left(\eta\frac{\partial u}{\partial y}\right)_{i+\myhalf , j-\myhalf}\right],\ ] ] with @xmath230 a similar construction is used for the mixed - derivative term , @xmath231_{i+\myhalf , j}=\d y^{-1}\left[\left(\eta\frac{\partial v}{\partial x}\right)_{i+\myhalf , j+\myhalf}-\left(\eta\frac{\partial v}{\partial x}\right)_{i+\myhalf , j-\myhalf}\right],\ ] ] with @xmath232 the stochastic stress tensor discretization is described in more detail in ref .
@xcite and applies in the present context as well .
for the low mach number equations , just as for the compressible equations , the symmetric form of the stochastic stress tensor must be used in order to ensure discrete fluctuation - dissipation balance between the viscous dissipation and stochastic forcing .
additionally , when @xmath10 is not spatially uniform the same interpolated viscosity @xmath233 as used in the viscous terms must be used when calculating the amplitude in the stochastic forcing @xmath234 at the edges ( nodes ) of the grid .
it is challenging to construct spatio - temporal discretizations that conserve the total mass while remaining consistent with the equation of state @xcite , as ensured in the continuum context by the constraint ( [ eq : div_v_constraint ] ) .
we demonstrate here how the special linear form of the constraint ( [ eq : eos_quasi_incomp ] ) can be exploited in the discrete context . following ref .
@xcite , we spatially discretize the advective terms in ( [ eq : rho1_eq ] ) using a centered ( skew - adjoint @xcite ) discretization , @xmath235_{i , j}=\delta x^{-1}\left[\left(\rho_{1}\right)_{i+\myhalf , j}u_{i+\myhalf , j}-\left(\rho_{1}\right)_{i-\myhalf , j}u_{i-\myhalf , j}\right]+\delta y^{-1}\left[\left(\rho_{1}\right)_{i , j+\myhalf}v_{i , j+\myhalf}-\left(\rho_{1}\right)_{i , j-\myhalf}v_{i , j-\myhalf}\right],\label{eq : rho_adv}\ ] ] and similarly for ( [ eq : rho_eq ] ) .
we would like this discrete advection to maintain the equation of state ( [ eq : eos_quasi_incomp ] ) at the discrete level , that is , maintain the constraint relating @xmath236 and @xmath237 in every cell @xmath238 . because the different dimensions are decoupled and the divergence is simply the sum of the one - dimensional difference operators , it is sufficient to consider ( [ eq : rho1_eq ] ) in one spatial dimension .
the method of lines discretization is given by the system of odes , one differential equation per cell @xmath239 , @xmath240,\ ] ] and similarly for @xmath241 .
as a shorthand , denote the quantity that appears in ( [ eq : eos_quasi_incomp ] ) with @xmath242 if we use the linear interpolation ( [ eq : rho_face ] ) to calculate face - centered densities , then because of the linearity of the eos the face - centered densities obey the eos if the cell - centered ones do , since @xmath243 .
the rate of change of @xmath244 in cell @xmath239 is @xmath245\\ & = & \left(\rho^{-1}\beta\right)\left(f_{i+\myhalf}-f_{i-\myhalf}\right)-\left(u_{i+\myhalf}-u_{i-\myhalf}\right)=0.\end{aligned}\ ] ] this simple calculation shows that the eos constraint @xmath246 is obeyed discretely in each cell at all times if it is initially satisfied and the velocities used to advect mass obey the discrete version of the constraint ( [ eq : div_v_constraint ] ) , @xmath247,\nonumber \end{aligned}\ ] ] in two dimensions .
our algorithm ensures that advective terms are always evaluated using a discrete velocity field that obeys this constraint .
this is accomplished by using a discrete projection operator , as we describe in the next section .
the spatial discretization of the advection terms in the momentum equation ( [ eq : momentum_eq ] ) is constructed using centered differences on the corresponding shifted ( staggered ) grid , as described in ref .
for example , for the @xmath201-component of momentum @xmath248 , @xmath249_{i+\myhalf , j}=\d x^{-1}\left[(m_{x}u)_{i+1,j}-(m_{x}u)_{i , j}\right]+\d y^{-1}\left[(m_{x}v)_{i+\myhalf , j+\myhalf}-(m_{x}v)_{i+\myhalf , j-\myhalf}\right],\label{eq : mom_adv}\ ] ] where simple averaging is used to interpolate momenta to the cell centers and edges ( nodes ) of the grid , for example , @xmath250 because of the linearity of the interpolation procedure , the interpolated discrete velocity used to advect @xmath202 obeys the constraint ( [ eq : div_v_mac ] ) on the shifted grid , with a right - hand side @xmath251 interpolated using the same arithmetic average used to interpolate the velocities .
in particular , in the incompressible case all variables , including momentum , are advected using a discretely divergence - free velocity , ensuring discrete fluctuation - dissipation balance @xcite . it is well - known that the centered discretization of advection we employ here is not robust for advection - dominated flows , and higher - order limiters and upwinding schemes are generally preferred in the deterministic setting @xcite .
however , these more robust advection schemes add artificial dissipation , which leads to a violation of discrete fluctuation - dissipation balance @xcite . in appendix [ appendixfiltering ]
we describe an alternative filtering procedure that can be used to handle strong advection while continuing to use centered differencing .
we now briefly discuss the spatial discretization of the affine operator @xmath109 defined by ( [ p_tilde_v ] ) , as used in our explicit temporal integrators .
the discrete projection takes a face - centered ( staggered ) discrete velocity field @xmath252 and a velocity divergence @xmath114 and projects @xmath122 onto the constraint ( [ eq : div_v_mac ] ) in a conservative manner . specifically , the projection consists of finding a cell - centered discrete scalar field @xmath253 such that @xmath254 where the gradient is discretized using centered differences , e.g. , @xmath255 the pressure correction @xmath253 is the solution to the variable - coefficient discrete poisson equation , @xmath256\nonumber \\
+ \frac{1}{\delta y}\left[\left(\frac{1}{\rho_{i , j+\myhalf}}\right)\left(\frac{\phi_{i , j+1}-\phi_{i , j}}{\delta y}\right)-\left(\frac{1}{\rho_{i , j-\myhalf}}\right)\left(\frac{\phi_{i , j}-\phi_{i , j-1}}{\delta y}\right)\right]\nonumber \\
= s_{i , j}-\left[\left(\frac{\tilde{u}_{i+\myhalf , j}-\tilde{u}_{i-\myhalf , j}}{\delta x}\right)+\left(\frac{\tilde{v}_{i , j+\myhalf}-\tilde{v}_{i , j-\myhalf}}{\delta y}\right)\right],\label{eq : discrete_poisson}\end{aligned}\ ] ] which can be solved efficiently using a standard multigrid approach @xcite .
the handling of different types of boundary conditions is relatively straightforward when a staggered grid is used and the physical boundaries are aligned with the cell boundaries for the scalar grid .
interpolation is not used to obtain values for faces , nodes or edges of the grid that lie on a physical boundary , since this would require `` ghost '' values at cell centers lying outside of the physical domain . instead ,
whenever a value of a physical variable is required at a face , node , or edge lying on a physical boundary , the boundary condition is used to obtain that value .
similarly , centered differences for the diffusive and viscous fluxes that require values outside of the physical domain are replaced by one - sided differences that only use values from the interior cell bordering the boundary and boundary values .
for example , if the concentration is specified at the face @xmath257 , the diffusive flux discretization ( [ eq : discrete_diff_flux ] ) is replaced with @xmath258 where @xmath259 is the specified boundary value , the density @xmath260 is obtained from @xmath259 using the eos constraint , and the diffusion coefficient @xmath215 is calculated at the specified values of concentration and density .
similar straightforward one - sided differencing is used for the viscous fluxes . as discussed in ref .
@xcite , the use of second - order one - sided differencing is not required to achieve global second - order accuracy , and would make the handling of the stochastic fluxes more complicated because it leads to a non - symmetric discrete laplacian . note that for the nonlinear low mach number equations our approach is subtly different from linearly extrapolating the value in the ghost cell @xmath261 .
namely , the extrapolated value might be unphysical , and it might not be possible to evaluate the eos or transport coefficients at the extrapolated concentration . for neumann - type or zero - flux boundary conditions ,
the corresponding diffusive flux is set to zero for any faces of the corresponding control volume that lie on physical boundaries , and values in cells outside of the physical domain are never required .
the corresponding handling of the stochastic fluxes is discussed in detail in ref .
@xcite .
the evaluation of advective fluxes for the scalars requires normal components of the velocity at the boundary . for faces of the grid that lie on a physical boundary ,
the normal component of the velocity is determined from the value of the diffusive mass flux at that face using ( [ eq : v_n_bc ] ) .
therefore , these velocities are not independent variables and are not solved for or modified by the projection @xmath109 . specifically , the discrete pressure @xmath253 is only defined at the cell centers in the interior of the grid , and the discrete poisson equation ( [ eq : discrete_poisson ] ) is only imposed on the interior faces of the grid .
therefore , no explicit boundary conditions for @xmath253 are required when the staggered grid is used , and the natural homogeneous neumann conditions are implied .
advective momentum fluxes are only evaluated on the interior faces and thus do not use any values outside of the physical domain .
by combining the spatial discretization described above with one of the temporal integators described in section [ sec : temporalintegration ] , we can obtain a finite - volume solver for the fluctuating low mach equations .
for the benefit of the reader , here we summarize our implementation of a single euler step ( [ eq : euler_step ] ) .
this forms the core procedure that the higher - order runge - kutta schemes employ several times during one time step . 1 .
generate the vectors of standard gaussian variates @xmath169 and @xmath170 .
2 . calculate diffusive and stochastic fluxes for @xmath36 using ( [ eq : discrete_diff_flux ] ) , @xmath262 3 .
solve the poisson problem ( [ eq : discrete_poisson ] ) with @xmath263 to obtain the velocity @xmath264 from @xmath265 using ( [ eq : u_minus_gradp ] ) , enforcing @xmath138 .
4 . calculate viscous and stochastic momentum fluxes using ( [ eq : div_viscous_stress ] ) , @xmath266^{n}+\grad\cdot\left[\m{\sigma}^{n}\left(\d t,\,\m w^{n}\right)\right].\ ] ] 5 .
calculate external forcing terms for the momentum equation , such as the contribution @xmath267 due to gravity .
calculate advective fluxes for mass and momentum using ( [ eq : rho_adv ] ) and ( [ eq : mom_adv ] ) .
update mass and momentum densities , including advective , diffusive , stochastic and external forcing terms , to obtain @xmath140 , @xmath141 and @xmath143 .
note that this update preserves the eos constraint as explained in section [ sub : advection ] .
we have tested and validated the accuracy of our methods and numerical implementation using a series of standard deterministic tests , as well as by examining the equilibrium spectrum of the concentration and velocity fluctuations @xcite .
the next two sections present further verification and validation in the context of nonequilibrium systems .
advection of concentration by thermal velocity fluctuations in the presence of large concentration gradients leads to the appearance of _ giant fluctuations _ of concentration , as has been studied theoretically and experimentally for more than a decade @xcite .
these giant fluctuations were previously simulated in the absence of gravity in three dimensions by some of us in ref .
@xcite , and good agreement was found with experimental results @xcite . in those previous studies
the incompressible equations were used , that is , it was assumed that concentration was a passively - advected scalar . however , it is more physically realistic to account for the fact that the properties of the fluid , notably the density and the transport coefficients , depend on the concentration . in ref .
@xcite a series of experiments were performed to study the temporal evolution of giant concentration fluctuations during the diffusive mixing of water and glycerol , starting with a glycerol mass fraction of @xmath268 in the bottom half of the experimental domain , and @xmath61 in the top half . because it is essentially impossible to analytically solve the full system of fluctuating equations in the presence of spatial inhomogeneity and nontrivial boundary conditions , the existing theoretical analysis of the diffusive mixing process @xcite makes a quasi - periodic constant - coefficient incompressible approximation . for simplicity , in this section
we focus on a time - independent problem and study the spectrum of steady - state concentration fluctuations in a mixture under gravity in the presence of a constant concentration gradient .
this extends the study reported in ref .
@xcite to account for the fact that the density , viscosity , and diffusion coefficient depend on the concentration . for simplicity
, we do two - dimensional simulations , since there are no qualitative differences between the spectrum of concentration fluctuations in two and three dimensions @xcite ( note , however , that in real space , unlike in fourier space , the effect of the fluctuations on the transport is very different in two and three dimensions ) . furthermore , in these simulations we do not include a stochastic flux in the concentration equation , i.e. , we set @xmath269 , so that all fluctuations in the concentration arise from being out of thermodynamic equilibrium . with this approximation
we do not need to model the chemical potential of the mixture and obtain @xmath270 .
this formulation is justified by the fact that it is known experimentally that the nonequilibrium fluctuations are much larger than the equilibrium ones for the conditions we consider @xcite . in the simple linearized theory presented in section [ sub : gianttheory ]
several approximations are made .
the first one is that a quasi - periodic approximation is used even though the actual system is not periodic in the @xmath203 direction .
this source of error has already been studied numerically in ref .
we also use a boussinesq approximation where it is assumed that @xmath271 and @xmath272 , where @xmath273 is a small density difference between the two fluids , @xmath274 , so that density is approximately constant and @xmath275 .
more precisely , in the boussinesq model the gravity term in the velocity equation only enters through the product @xmath276 so the approximation consists of taking the limit @xmath277 and @xmath278 while keeping the product @xmath276 fixed .
the final approximation made in the simple theory is that the transport coefficients , i.e. , the viscosity and diffusion coefficients , are assumed to be constant .
here we evaluate the validity of the constant - coefficient constant - density approximation ( @xmath279 and @xmath12 constant , @xmath277 ) , as well as the constant - density ( boussinesq ) approximation alone ( @xmath33 constant , @xmath277 , but variable @xmath280 ) , by comparing with the solution to the complete low mach number equations ( @xmath281 and @xmath282 variable ) .
we base our parameters on the experimental studies of diffusive mixing in a water - glycerol mixture , as reported in ref .
the physical domain is @xmath283 discretized on a uniform @xmath284 two dimensional grid , with a thickness of @xmath285 along the @xmath286 direction .
gravity is applied in the negative @xmath203 ( vertical ) direction .
reservoir boundary conditions ( [ eq : v_n_bc ] ) are applied in the @xmath203-direction and periodic boundary conditions in the @xmath201-direction .
we set the concentration to @xmath268 on the bottom boundary and @xmath61 on the top boundary , and apply no - slip boundary conditions for the velocity at both boundaries .
the initial condition is @xmath287 , which is close to the deterministic steady - state profile .
a very good fit to the experimental equation of state ( dependence of density on concentration at standard temperature and pressure ) over the whole range of concentrations of interest is provided by the eos ( [ eq : eos_quasi_incomp ] ) with the density of water set to @xmath288 and the density of glycerol set to @xmath289 . in these simulations
the magnitude of the velocity fluctuations is very small and we did not use filtering ( see appendix [ appendixfiltering ] ) . experimentally , the dependence of viscosity on glycerol mass fraction has been fit to an exponential function @xcite , which we approximate with a quadratic function over the range of concentrations of interest , @xmath290 where @xmath291 and experimental measurements estimate @xmath292 .
the diffusion coefficient dependence on the concentration has been studied experimentally @xcite , but is in fact strongly affected by thermal fluctuations and spatial confinement @xcite .
we approximate the dependence assuming a stokes - einstein relation @xcite , which is in reasonable agreement with the experimental results in ref .
@xcite over the range of concentrations of interest here , @xmath293 where experimental estimates @xcite for water - glycerol mixtures give @xmath294 , with a schmidt number @xmath295 .
this very large separation of scales between mass and momentum diffusion is not feasible to simulate with our explicit temporal integration methods . referring back to the simplified theory ( [ eq : s_c_c ] ) , which in this case can be simplified further to @xmath296 we see that for @xmath297 the shape of the spectrum of the steady - state concentration fluctuations , and in particular , the cutoff wavenumber due to gravity , is determined from the product @xmath298 and not @xmath12 and @xmath10 individually .
therefore , as also done in ref .
@xcite , we choose @xmath299 and @xmath300 so that @xmath301 is kept at the physical value of @xmath302 but the schmidt number is reduced by two orders of magnitude , @xmath303 , where @xmath304 is an estimate of the average concentration .
the condition @xmath305 and @xmath306 gives our simulation parameters @xmath307 and @xmath308 .
the physical value for gravity is @xmath309 and the solutal expansion coefficient @xmath310 follows from @xmath58 and @xmath59 . when employing the boussinesq approximation , in which gravity only enters through the product @xmath311 we set @xmath312 and @xmath313 so that @xmath314 and increase gravity by the corresponding factor to @xmath315 in order to keep @xmath276 fixed at the physical value .
we also performed simulations with a weaker gravity , @xmath316 , which enhances the nonequilibrium fluctuations , as well as no gravity , which makes the fluctuations truly giant @xcite .
we employ the explicit midpoint temporal integrator ( which we recall is third - order accurate for static covariances ) and set @xmath317 , which results in a diffusive courant number @xmath318 .
we skip the first 50,000 time steps ( about 5 diffusion crossing times ) and then collect samples from the subsequent 50,000 time steps .
we repeat this eight times to increase the statistical accuracy and estimate error bars . to compare to the theory ( [ eq : s_c_c ] ) , we set the concentration gradient to @xmath319 and evaluate @xmath320 at @xmath321 from the equation of state .
when computing the theory , we account for errors in the discrete approximation to the continuum laplacian by using the effective wavenumber @xmath322 instead of the actual discrete wavenumber @xmath323 @xcite .
comparison between the simple theory ( [ eq : s_c_c_simp ] ) ( lines ) and numerical results ( symbols ) .
results are shown for standard gravity @xmath324 ( the cutoff wavenumber @xmath325 ) , for the complete variable - coefficient variable - density low mach model ( green upward triangles ) and the constant - coefficient constant - density approximation ( red squares ) .
also shown are results for a weaker gravity , @xmath326 ( the cutoff wavenumber @xmath327 ) , for the complete low mach model ( magenta pluses ) and the constant - coefficient constant - density approximation ( cyan stars ) . for comparison ,
results for @xmath326 with variable viscosity @xmath328 but constant diffusion coefficient @xmath329 are also shown , for variable density ( orange downward triangles ) and the constant - density ( boussinesq ) approximation ( indigo right - facing triangles ) . finally , results for no gravity are shown in the constant - coefficient approximation ( black circles).,scaledwidth=75.0% ] the results for the static spectrum of concentration fluctuations @xmath330 as a function of the modified wavenumber @xmath323 ( [ eq : modified_kx ] ) are shown in fig .
[ fig : constantgradient ] . when there is no gravity , we see the characteristic giant fluctuation power - law spectrum of the fluctuations , modulated at small wavenumbers due to the presence of the physical boundaries @xcite . when gravity is present , fluctuations at wavenumber below the cutoff @xmath331^{1/4}$ ] are suppressed .
if we use a constant - coefficient approximation , in which we reduce @xmath314 so that @xmath332 and also fix the transport coefficients at @xmath333 and @xmath334 , we observe good agreement with the quasi - periodic theory ( [ eq : s_c_c_simp ] ) .
when we make the transport coefficients dependent on the concentration as in ( [ eq : nu_c],[eq : chi_c ] ) , we observe a rather small change in the spectrum .
this is perhaps not unexpected because the simplified theory ( [ eq : s_c_c_simp ] ) shows that only the product @xmath298 , and not @xmath12 and @xmath10 individually , matters . since we used the stokes - einstein relation @xmath335 to select the concentration dependence of the diffusion coefficient ,
the value of @xmath298 is constant throughout the physical domain . for comparison , in fig .
[ fig : constantgradient ] we show results from a simulation where we keep the concentration dependence of the viscosity ( [ eq : nu_c ] ) but set the diffusion coefficient to a constant value , @xmath329 , and we observe a more significant change in the spectrum .
further employing the boussinesq approximation makes little difference showing that the primary effect here comes from the dependence of the transport coefficients on concentration .
this shows that under the sort of parameters present in the experiments on diffusive mixing in water - glycerol mixture , it is reasonable to make the boussinesq incompressible approximation ; however , the spatial dependence of the viscosity and diffusion coefficient can not in general be ignored if quantitative agreement is desired . in particular , time - dependent quantities such as dynamic spectra @xcite depend on the individual values of @xmath12 and @xmath10 and not just their product , and are thus expected to be more sensitive to the details of their concentration dependence .
even though the constant - coefficient approximation gives qualitatively the correct shape and a better choice of the constant transport coefficients may improve its accuracy , there is no obvious or simple procedure to _ a priori _ estimate what parameters should be used ( but see @xcite for a proposal to average the constant - coefficient theory over the domain ) .
a direct comparison with experimental results is not possible until multiscale temporal integrators capable of handling the extreme separation of time scales between mass and momentum diffusion are developed . at present
this has only been accomplished in the constant - coefficient incompressible limit ( @xmath43 ) @xcite , and it remains a significant challenge to accomplish the same for the complete low mach number system .
in this section we study the appearance of giant fluctuations during _ time - dependent _ diffusive mixing . as a validation of the low mach number fluctuating equations and our algorithm , we perform simulations of diffusive mixing of two fluids of different densities in two dimensions .
we find excellent agreement between the results of low mach number ( continuum ) simulations and hard - disk molecular dynamics ( particle ) simulations .
this nontrivial test clearly demonstrates the usefulness of low mach number models as a coarse - grained mesoscopic model for problems where sound waves can be neglected .
our simulation setup is illustrated in fig .
[ fig : mixingillustrationmass4 ] .
we consider a periodic square box of length @xmath336 along both the @xmath201 ( horizontal ) and @xmath203 ( vertical ) directions , and initially place all of the fluid of species one ( colored red ) in the middle third of the domain , i.e. , we set @xmath60 for @xmath337 , and @xmath61 otherwise , as shown in the top left panel of the figure . the two fluids mix diffusively and at the end of the simulation the concentration field shows a _ rough diffusive interface _ as confirmed by molecular dynamics simulations shown in the top right panel of the figure .
the deterministic equations of diffusive mixing reduce to a one dimensional model due to the translational symmetry along the @xmath201 axes , and would yield a _
flat _ diffusive interface as illustrated in the bottom left panel of the figure .
however , fluctuating hydrodynamics correctly reproduces the interface roughness , as illustrated in the bottom right panel of the figure and demonstrated quantitatively below .
diffusive mixing between two fluids of unequal densities , @xmath338=4 , with coloring based on concentration , red for the pure first component , @xmath60 , and blue for the pure second component , @xmath61 .
a smoothed shading is used for the coloring to eliminate visual discretization artifacts .
the simulation domain is periodic and contains @xmath339 hydrodynamic ( finite volume ) cells .
the top left panel shows the initial configuration , which is the same for all simulations reported here .
the top right panel shows the final configuration at time @xmath340 as obtained using molecular dynamics .
the bottom left panel shows the final configuration obtained using deterministic hydrodynamics , while the right panel shows the final configuration obtained using fluctuating hydrodynamics.,title="fig:",scaledwidth=40.0%]diffusive mixing between two fluids of unequal densities , @xmath338=4 , with coloring based on concentration , red for the pure first component , @xmath60 , and blue for the pure second component , @xmath61 .
a smoothed shading is used for the coloring to eliminate visual discretization artifacts .
the simulation domain is periodic and contains @xmath339 hydrodynamic ( finite volume ) cells .
the top left panel shows the initial configuration , which is the same for all simulations reported here .
the top right panel shows the final configuration at time @xmath340 as obtained using molecular dynamics .
the bottom left panel shows the final configuration obtained using deterministic hydrodynamics , while the right panel shows the final configuration obtained using fluctuating hydrodynamics.,title="fig:",scaledwidth=40.0% ] diffusive mixing between two fluids of unequal densities , @xmath338=4 , with coloring based on concentration , red for the pure first component , @xmath60 , and blue for the pure second component , @xmath61 .
a smoothed shading is used for the coloring to eliminate visual discretization artifacts .
the simulation domain is periodic and contains @xmath339 hydrodynamic ( finite volume ) cells .
the top left panel shows the initial configuration , which is the same for all simulations reported here .
the top right panel shows the final configuration at time @xmath340 as obtained using molecular dynamics .
the bottom left panel shows the final configuration obtained using deterministic hydrodynamics , while the right panel shows the final configuration obtained using fluctuating hydrodynamics.,title="fig:",scaledwidth=40.0%]diffusive mixing between two fluids of unequal densities , @xmath338=4 , with coloring based on concentration , red for the pure first component , @xmath60 , and blue for the pure second component , @xmath61 .
a smoothed shading is used for the coloring to eliminate visual discretization artifacts .
the simulation domain is periodic and contains @xmath339 hydrodynamic ( finite volume ) cells .
the top left panel shows the initial configuration , which is the same for all simulations reported here .
the top right panel shows the final configuration at time @xmath340 as obtained using molecular dynamics .
the bottom left panel shows the final configuration obtained using deterministic hydrodynamics , while the right panel shows the final configuration obtained using fluctuating hydrodynamics.,title="fig:",scaledwidth=40.0% ] we consider here a binary hard - disk mixture in two dimensions .
we use arbitrary ( molecular ) units of length , time and mass for convenience .
all hard disks had a diameter @xmath341 in arbitrary units , and we set the temperature at @xmath342 .
the molecular mass for the first fluid component was fixed at @xmath343 , and for the second component at @xmath344 . for mass ratio @xmath345 , the two types of disks are mechanically - identical and
therefore the species label is simply a red - blue coloring of the particles . in this case @xmath346 and the low mach number equations reduce to the incompressible equations of fluctuating hydrodynamics with a passively - advected concentration field . for the case of unequal particle masses ,
mechanical equilibrium is obtained if the pressures in the two fluid components are the same .
it is well - known from statistical mechanics that for hard disks or hard spheres the pressure is @xmath347 where @xmath348 is the number density , and @xmath349 is a prefactor that only depends on the packing fraction @xmath350 and not on the molecular mass .
therefore , for a mixture of disks or spheres with equal diameters , at constant pressure , the number density and the packing fraction @xmath253 are constant independent of the composition .
the equation of state at constant pressure and temperature is therefore @xmath351 which is exactly of the form ( [ eq : eos_quasi_incomp ] ) with @xmath352 and @xmath67 .
the chemical potential of such a mixture has the same concentration dependence as a low - density gas mixture @xcite , @xmath353.\ ] ] in order to validate the predictions of our low mach number model , we performed hard disk molecular dynamics ( hdmd ) simulations of diffusive mixing using a modification of the public - domain code developed by the authors of ref .
we used a packing fraction of @xmath354 for all simulations reported here .
this packing fraction is close to the freezing transition point but is known to be safely in the ( dense ) gas phase ( there is no liquid phase for a hard - disk fluid ) .
the initial particle positions were generated using a nonequilibrium molecular dynamics simulation as in the hard - particle packing algorithm described in ref .
after the initial configuration was generated the disks were assigned a species according to their @xmath203 coordinate , and the mixing simulation performed using event - driven molecular dynamics . in order to convert the particle data to hydrodynamic data comparable to that generated by the fluctuating hydrodynamics simulations
, we employed a grid of @xmath355 hydrodynamic cells that were each a square of linear dimension @xmath356 . at the chosen packing fraction @xmath354 this corresponds to about @xmath357 disks per hydrodynamic cell , which is deemed a reasonable level of coarse - graining for the equations of fluctuating hydrodynamics to be a reasonably - accurate model , while still keeping the computational demands of the simulations manageable .
we performed hdmd simulations for systems of size @xmath358 and @xmath359 cells , and simulated the mixing process to a final simulation time of @xmath340 units .
the largest system simulated had about @xmath360 million disks ( each simulation took about 5 days of cpu time ) , which is well into the `` hydrodynamic '' rather than `` molecular '' scale .
every @xmath361 units of time , particle data was converted to hydrodynamic data for the purposes of analysis and comparison to hydrodynamic calculations . there is not a unique way of coarse - graining particle data to hydrodynamic data @xcite ; however , we believe that the large - scale ( giant ) concentration fluctuations studied here are _ not _ affected by the particular choice .
we therefore used a simple method consistent with the philosophy of finite - volume conservative discretizations .
specifically , we coarse - grained the particle information by sorting the particles into hydrodynamic cells based on the position of their centroid , as if they were point particles .
we then calculated @xmath36 and @xmath196 in each cell based on the total mass of each species contained inside the given cell .
since all particles have equal diameter other definitions that take into account the particle shape and size give similar results .
we now turn to hydrodynamic simulations of the diffusive mixing of hard disks .
our hydrodynamic calculations use the same grid of cells used to convert particle to hydrodynamic data . the only input required for the hydrodynamic calculations , in addition to those provided by equilibrium statistical mechanics , are the transport coefficients of the fluid as a function of concentration , specifically , the shear viscosity @xmath10 and the diffusion coefficient @xmath12 . the values for the transport coefficients used in the spatio - temporal discretization , as explained in refs .
@xcite and detailed in appendix [ sec : transportmd ] , are not material constants independent of the discretization .
rather , they are _ bare _ transport values @xmath362 and @xmath299 measured at the length scales of the grid size .
we assumed that the bare transport coefficients obey the same scaling with the mass ratio @xmath363 as predicted by enskog kinetic theory ( [ eq : eta_r_scaling],[eq : chi_r_scaling ] ) .
as explained in appendix [ sec : transportmd ] , theoretical arguments and molecular dynamics results suggest that renormalization effects for viscosity are small and can be safely neglected .
we have therefore fixed the viscosity in the hydrodynamic calculations based on the molecular dynamics estimate @xmath364 for the pure fluid with molecular mass @xmath365 ( see section [ sub : viscosity ] ) . however , the bare diffusion coefficient is strongly dependent on the size of the hydrodynamic cells ( held fixed in our calculations at @xmath366 ) , and on whether filtering ( see appendix [ appendixfiltering ] ) is used
. therefore , the value of @xmath299 needs to be adjusted based on the spatial discretization , in such a way as to match the behavior of the molecular dynamics simulations at length scales much larger than the grid spacing .
we describe the exact procedure we used to accomplish this in section [ sub : diffusioncoeffient ] .
the time step in our explicit algorithm is limited by the viscous cfl number @xmath367 . since
the hydrodynamic calculations are much faster compared to the particle simulations , we used the more expensive rk3 temporal integrator with a relatively small time step @xmath368 , corresponding to @xmath369 for @xmath60 . for @xmath345 and @xmath358 we employed a larger time step , @xmath370 ( @xmath371 ) , with no measurable temporal discretization artifacts for the quantities studied here . we are therefore confident that the discretization errors in this study are dominated by spatial discretization artifacts . in future work
we will explore semi - implicit discretizations and study the effect of taking larger time steps on temporal accuracy .
note that at these parameters for @xmath60 the isothermal speed of sound is @xmath372 so that a compressible scheme would require a time step on the order of @xmath373 ( corresponding to advective cfl of about a half ) .
by contrast , the explicit low mach number algorithm is stable for @xmath374 .
this modest gain is due to the small hydrodynamic cell we use here in order to compare to molecular dynamics . for mesoscopic
hydrodynamic cells the gain in time step size afforded by the low mach formulation will be several orders of magnitude larger . for mass ratio @xmath345 and @xmath375 , the hydrodynamic calculations were initialized using statistically identical configurations as would be obtained by coarse - graining the initial particle configuration .
this implies a sharp , step - like jump in concentration at @xmath376 and @xmath377 .
since our spatio - temporal discretization is not strictly monotonicity - preserving , such sharp concentration gradients combined with a small diffusion coefficient @xmath299 lead to a large cell peclet number .
this may in turn lead to large deviations of concentration outside of the allowed interval @xmath68 for larger mass ratios .
therefore , for @xmath378 we smoothed the initial condition slightly so that the sharp jump in concentration is spread over a few cells , and also employed a @xmath379 point filter for the advection velocity ( @xmath380 , see appendix [ appendixfiltering ] ) .
we verified that for @xmath375 using filtering only affects the large wavenumbers and does not appear to affect the small wavenumbers we study here , provided the bare diffusion coefficient @xmath299 is adjusted based on the specific filtering width @xmath381 . in order to compare the molecular dynamics and the hydrodynamic simulations we calculated several statistical quantities : 1 .
the averages of @xmath36 along the directions perpendicular to the concentration gradient , @xmath382 where the integral is discretized as a direct sum over the hydrodynamic cells .
note that it is statistically better to use conserved quantities for such macroscopic averages than to use non - conserved variables such as concentration @xcite .
the spectrum of the concentration averaged along the direction of the gradient by computing the average @xmath383 and then taking the discrete fourier transform .
intuitively , @xmath384 is a measure of the thickness of the red strip in fig .
[ fig : mixingillustrationmass4 ] , and corresponds closely to what is measured in light scattering and shadowgraphy experiments @xcite .
the discrete fourier spectrum of the @xmath203-coordinate of the `` center - of - mass '' of concentration along the direction perpendicular to the gradient , @xmath385 intuitively , @xmath386 is a measure of the height of the centerline of the red strip in fig .
[ fig : mixingillustrationmass4 ] .
all quantities were sampled at certain pre - specified time points in a number of statistically - independent simulations @xmath387 and then means and standard deviations calculated from the @xmath387 data points . for system of size @xmath358 cells we used @xmath388 simulations , and for systems of size @xmath359 we used @xmath389 simulations . by far the majority of the computational cost was in performing the hdmd simulations .
once @xmath299 and @xmath390 were estimated based on simulations of a constant - density ( @xmath345 ) fluid ( see section [ sub : diffusioncoeffient ] ) , kinetic theory ( [ eq : eta_r_scaling],[eq : chi_r_scaling ] ) can be used to estimate them for different density ratios . in fig .
[ fig : spreading_64x64 ] we show @xmath391 for mass ratio @xmath375 , showing good agreement between hdmd and hydrodynamics , especially when fluctuations are accounted for . for @xmath378
a direct comparison is difficult because the initial condition was slightly different in the hydrodynamic simulations due to the need to smooth the sharp concentration gradient for numerical reasons , as explained earlier .
this difference strongly affects the shape of @xmath391 at early times , however , it does not significantly modify the roughness of the interface , which we study next .
the most interesting contribution of fluctuations to the diffusive mixing process is the appearance of giant concentration fluctuations in the presence of large concentration gradients , as evidenced in the roughness of the interface between the two fluids during the early stages of the mixing in fig .
[ fig : mixingillustrationmass4 ] . in order to quantify this interface roughness we used the one - dimensional power spectra @xmath392
note that here we do not correct the discrete wavenumber for the spatial discretization artifacts and continue to use @xmath323 instead of @xmath393 .
discrete spatial spectrum of the interface fluctuations for @xmath345 and @xmath359 ( averaged over 32 simulations ) at several points in time ( drawn with different colors , as indicated in the legend ) , for fluctuating hydrodynamics ( fh , squares with error bars ) and hdmd ( circles , error bars comparable to those for squares ) .
note that the largest wavenumber supported by the grid is @xmath394 .
the larger wavenumbers are however dominated by spatial truncation errors and the filter employed ( if any ) and we do not show them here .
( left panel ) spectrum @xmath395 of the vertically - averaged concentration .
( right panel ) spectrum @xmath396 of the position of the vertical `` center - of - mass '' of concentration.,title="fig:",scaledwidth=49.0%]discrete spatial spectrum of the interface fluctuations for @xmath345 and @xmath359 ( averaged over 32 simulations ) at several points in time ( drawn with different colors , as indicated in the legend ) , for fluctuating hydrodynamics ( fh , squares with error bars ) and hdmd ( circles , error bars comparable to those for squares ) .
note that the largest wavenumber supported by the grid is @xmath394 .
the larger wavenumbers are however dominated by spatial truncation errors and the filter employed ( if any ) and we do not show them here .
( left panel ) spectrum @xmath395 of the vertically - averaged concentration .
( right panel ) spectrum @xmath396 of the position of the vertical `` center - of - mass '' of concentration.,title="fig:",scaledwidth=49.0% ] the temporal evolution of the spectra @xmath397 and @xmath398 is shown in fig .
[ fig : s_k_mass_1 ] for mass ratio @xmath345 , and in fig .
[ fig : s_k_mass_4 ] for mass ratio @xmath378 , for both hdmd and low mach number fluctuating hydrodynamics ( note that deterministic hydrodynamics would give identically zero for any spectral quantity ) .
we observe an excellent agreement between the two , including the correct initial evolution of the interface fluctuations .
same as fig .
[ fig : s_k_mass_1 ] but for density ratio @xmath378.,title="fig:",scaledwidth=49.0%]same as fig .
[ fig : s_k_mass_1 ] but for density ratio @xmath378.,title="fig:",scaledwidth=49.0% ] note that for a finite system , eventually complete mixing will take place and the concentration fluctuations will have to revert to their equilibrium spectrum , which is flat in fourier space instead of the power - law behavior seen out of equilibrium . in fig .
[ fig : s_k_mass_1_long ] we show results for mixing up to a time @xmath399 ( this is 128 times longer than those described above ) .
these long simulations are only feasible for the fluctuating hydrodynamics code , and employ a somewhat larger time step @xmath370 .
the results clearly show that at late times the spectrum of the fluctuations reverts to the equilibrium one ; however , this takes some time even after the mixing is essentially complete .
linearized incompressible fluctuating hydrodynamics @xcite predicts that at steady state the spectrum of nonequilibrium concentration fluctuations is a power law with exponent @xmath400 , @xmath401 the dynamically - evolving spectra in the right panel of fig .
[ fig : s_k_mass_1_long ] show approximately such power - law behavior for intermediate times and wavenumbers . mixing to a time 128 times longer than previous results , with results reported at time intervals @xmath402 for @xmath403 .
these long simulations are only feasible for the fluctuating hydrodynamics code , and employ a somewhat larger time step @xmath370 .
( _ left _ ) horizontally - averaged @xmath36 , as shown for the shorter runs in the left panel of fig .
[ fig : spreading_64x64 ] .
( _ right _ ) the spectrum of interface fluctuations @xmath395 , as shown in the left panels of figs .
[ fig : s_k_mass_1 ] and [ fig : s_k_mass_4 ] for the shorter runs .
the theoretical estimates for the spectrum of equilibrium fluctuations , which is independent of wavenumber , is also shown .
we also indicate the theoretical prediction for the power - law of the spectrum of steady - state nonequilibrium fluctuations under an applied concentration gradient , @xmath404,title="fig:",scaledwidth=49.0%]mixing to a time 128 times longer than previous results , with results reported at time intervals @xmath402 for @xmath403 .
these long simulations are only feasible for the fluctuating hydrodynamics code , and employ a somewhat larger time step @xmath370 .
( _ left _ ) horizontally - averaged @xmath36 , as shown for the shorter runs in the left panel of fig .
[ fig : spreading_64x64 ] .
( _ right _ ) the spectrum of interface fluctuations @xmath395 , as shown in the left panels of figs .
[ fig : s_k_mass_1 ] and [ fig : s_k_mass_4 ] for the shorter runs .
the theoretical estimates for the spectrum of equilibrium fluctuations , which is independent of wavenumber , is also shown .
we also indicate the theoretical prediction for the power - law of the spectrum of steady - state nonequilibrium fluctuations under an applied concentration gradient , @xmath404,title="fig:",scaledwidth=49.0% ] in order to illustrate the appearance of giant fluctuations in three dimensions we performed simulations of mixing in a mixture of hard spheres with equal diameters , @xmath341 , and mass ratio @xmath378 . the packing density was chosen to be @xmath405 , which corresponds to a very dense gas , but is still well below the freezing point @xmath406 . for the hydrodynamic simulations we used cubic cells of dimension @xmath407 , which corresponds to about @xmath408 particles per hydrodynamic cell on average . in fig .
[ fig : s_k_mass_4_3d ] we show results from a single simulation with a grid of size @xmath409 cells , which would correspond to about @xmath410 particles .
this makes molecular dynamics simulations infeasible , and makes hydrodynamic calculations an invaluable tool in studying the mixing process at these mesoscopic scales . in the hydrodynamic simulations we used bare transport coefficient values based on enskog kinetic theory for the hard - sphere fluid @xcite . for the single - component fluid with molecular mass @xmath365
, this theory gives @xmath411 and @xmath412 , which corresponds to a bare schmidt number @xmath413 .
we employed the same model dependence of bare transport coefficients on concentration as for hard disks , see eqs .
( [ eq : eta_r_scaling],[eq : chi_r_scaling ] ) .
the time step was set at @xmath414 ( corresponding to viscous cfl number @xmath415 ) . in three dimensions ,
the cell peclet number is reduced with decreasing @xmath128 and we did not find it necessary to employ any filtering .
diffusive mixing in three dimensions similar to that illustrated in fig .
[ fig : mixingillustrationmass4 ] for two dimensions .
parameters are based on enskog kinetic theory for a hard - sphere fluid at packing fraction @xmath405 , and there is no gravity .
the mixing starts with the top half being one species and the bottom half another species , with density ratio @xmath378 , and concentration is kept fixed at the top and bottom boundaries while the side boundaries are periodic . a snapshot taken at time @xmath416 is shown .
( _ top panel _ )
the side panes show two dimensional slices for the concentration @xmath34 . the approximated contour surface @xmath417 is shown with color based on surface height to illustrate the rough diffusive interface .
( bottom left panel ) similar as top panel but bottom pane shows vertically - averaged concentration @xmath418 , illustrating the giant concentration fluctuations .
( bottom right panel ) the fourier spectrum @xmath419 of @xmath384 .
the color axes is logarithmic and clearly shows the appearance of large scale ( small wavenumber ) fluctuations , as also seen in fig .
[ fig : s_k_mass_4 ] in two dimensions.,scaledwidth=75.0% ] diffusive mixing in three dimensions similar to that illustrated in fig .
[ fig : mixingillustrationmass4 ] for two dimensions .
parameters are based on enskog kinetic theory for a hard - sphere fluid at packing fraction @xmath405 , and there is no gravity .
the mixing starts with the top half being one species and the bottom half another species , with density ratio @xmath378 , and concentration is kept fixed at the top and bottom boundaries while the side boundaries are periodic . a snapshot taken at time
@xmath416 is shown .
( _ top panel _ )
the side panes show two dimensional slices for the concentration @xmath34 . the approximated contour surface @xmath417 is shown with color based on surface height to illustrate the rough diffusive interface .
( bottom left panel ) similar as top panel but bottom pane shows vertically - averaged concentration @xmath418 , illustrating the giant concentration fluctuations .
( bottom right panel ) the fourier spectrum @xmath419 of @xmath384 .
the color axes is logarithmic and clearly shows the appearance of large scale ( small wavenumber ) fluctuations , as also seen in fig .
[ fig : s_k_mass_4 ] in two dimensions.,title="fig:",scaledwidth=49.0%]diffusive mixing in three dimensions similar to that illustrated in fig .
[ fig : mixingillustrationmass4 ] for two dimensions .
parameters are based on enskog kinetic theory for a hard - sphere fluid at packing fraction @xmath405 , and there is no gravity .
the mixing starts with the top half being one species and the bottom half another species , with density ratio @xmath378 , and concentration is kept fixed at the top and bottom boundaries while the side boundaries are periodic . a snapshot taken at time
@xmath416 is shown .
( _ top panel _ )
the side panes show two dimensional slices for the concentration @xmath34 . the approximated contour surface @xmath417 is shown with color based on surface height to illustrate the rough diffusive interface .
( bottom left panel ) similar as top panel but bottom pane shows vertically - averaged concentration @xmath418 , illustrating the giant concentration fluctuations .
( bottom right panel ) the fourier spectrum @xmath419 of @xmath384 .
the color axes is logarithmic and clearly shows the appearance of large scale ( small wavenumber ) fluctuations , as also seen in fig .
[ fig : s_k_mass_4 ] in two dimensions.,title="fig:",scaledwidth=49.0% ] instead of the fully periodic domain used in the two dimensional hard - disk simulations , here we employ the fixed - concentration boundary conditions ( [ eq : v_n_bc ] ) and set @xmath420 at the bottom and @xmath421 at the top boundary .
this emulates the sort of `` open '' or `` reservoir '' boundaries @xcite that mimic conditions in experimental studies of diffusive mixing @xcite .
the initial condition is a fully phase - separated mixture with @xmath60 for @xmath422 , and @xmath61 otherwise . as the mixing process continues the diffusive interface roughens and giant concentrations
appear , as illustrated in fig .
[ fig : s_k_mass_4_3d ] and also observed experimentally in water - glycerol mixtures in ref .
@xcite . in three dimensions , however , the diffusive interface roughness is much smaller than in two dimensions , being on the order of only 20 molecular diameters for the snapshot shown in the figure .
this illustrates the importance of dimensionality when including thermal fluctuations .
in particular , unlike in deterministic fluid dynamics , in fluctuating hydrodynamics one can not simply eliminate dimensions from consideration even in simple geometries .
approximate theory based on the boussinesq approximation and linearization of the equations of fluctuating hydrodynamics has been developed in ref .
@xcite and applied in the analysis of experimental results on mixing in a water - glycerol mixture in the presence of gravity @xcite .
the simulations reported here do not make the sort of approximations necessary in analytical theories and can in principle be used to study the mixing process quantitatively .
however , it is important to emphasize that in realistic liquids , such as a water - glycerol mixture , the schmidt number is on the order of a thousand .
this makes explicit time stepping schemes that fully resolve the dynamics of the velocity fluctuations infeasible . in future work
we will consider semi - implicit type stepping methods that relax the severe time stepping restrictions present in the explicit schemes considered here .
the behavior of fluids is strongly affected by thermal fluctuations at scales from the microscopic to the macroscopic .
fluctuating hydrodynamics is a powerful coarse - grained model for fluid dynamics at mesoscopic and macroscopic scales , at both a theoretical and a computational level .
theoretical calculations are rather complicated in the presence of realistic spatial inhomogeneities and nontrivial boundary conditions . in numerical simulations , those effects can readily be handled , however , the large separation of time scales between different physical processes poses a fundamental difficulty .
compressible fluctuating hydrodynamics bridges the gap between molecular and hydrodynamic scales . at spatial scales
not much larger than molecular , sound and momentum and heat diffusion occur at comparable time scales in both gases and liquids . at mesoscopic and
larger length scales , fast pressure fluctuations due to thermally - actuated sound waves are much faster than diffusive processes .
it is therefore necessary to eliminate sound modes from the compressible equations . in the deterministic context
this is accomplished using low mach number asymptotic expansion . for homogeneous simple fluids or mixtures of dynamically - identical fluids the zeroth order low mach equations
are the well - known incompressible navier - stokes equations , in which pressure is a lagrange multiplier enforcing a divergence - free velocity field . in mixtures of dissimilar fluids , local changes in composition and temperature cause local expansion and contraction of the fluid and thus a nonzero velocity divergence . in this paper
we proposed low mach number fluctuating equations for isothermal binary mixtures of incompressible fluids with different density , or a mixture of low - density gases with different molecular masses .
these equations are a straightforward generalization of the widely - used incompressible fluctuating navier - stokes equations . in the low mach number equations
the incompressibility constraint @xmath423 is replaced by @xmath424 , which ensures that compositional changes are accompanied by density changes in agreement with the fluid equation of state ( eos ) at constant pressure and temperature .
this seemingly simple generalization poses many non - trivial analytical and numerical challenges , some of which we addressed in this paper . at the analytical level
the low mach number fluctuating equations are different from the incompressible equations because the velocity divergence is directly coupled to the time derivative of the concentration fluctuations .
this means that at thermodynamic equilibrium the velocity is not only white in space , a well - known difficulty with the standard equations of fluctuating hydrodynamics , but is also white in time , adding a novel type of difficulty that has not heretofore been recognized .
the unphysically fast fluctuations in velocity are caused by the unphysical assumption of infinite separation of time scales between the sound and the diffusive modes .
this unphysical assumption also underlies the incompressible fluctuating navier - stokes equations , however , in the incompressible limit @xmath277 the problem is not apparent because the component of velocity that is white in time disappears .
here we analyzed the low mach equations at the linearized level , and showed that they reproduce the slow diffusive fluctuations in the full compressible equations , while eliminating the fast pressure fluctuations . at the formal level , we suggest that a generalized hodge decomposition can be used to separate the vortical ( solenoidal ) modes of velocity as the independently fluctuating variable , coupled with a gauge formulation used to treat the divergence constraint . such nonlinear analysis is deferred for future research , and here we relied on the fact that the temporal discretization regularizes the short - time dynamics at time scales faster than the time step size @xmath126 . at the numerical level ,
the low mach number equations pose several distinct challenges .
the first challenge is to construct conservative spatial discretizations in which density is advected in a locally - conservative manner while still maintaining the equation of state constraint relating the local densities and composition .
we accomplish this here by using a specially - chosen model eos that is linear yet still rather versatile in practice , and by advecting densities using a velocity that obeys a discrete divergence constraint .
we note that for this simplified case , the system can be modeled using only the concentration to describe the thermodynamic state .
however , for more general low mach number models maintaining a full thermodynamic representation of the state independent of the constraint leads to more robust numerics . as in incompressible hydrodynamics , enforcing this constraint requires a poisson pressure solver that dominates the computational cost of the algorithm .
a second challenge is to construct temporal integrators that are at least second - order in time .
we accomplish this here by formally introducing an unconstrained gauge formulation of the equations , while at the same time taking advantage of the gauge degree of freedom to avoid ever explicitly dealing with the gauge variable .
the present temporal discretizations are purely explicit and are similar in spirit to an explicit projection method . a third and remaining challenge is to design efficient temporal integrators that handle momentum diffusion , the second - fastest physical process , semi - implicitly .
this poses well - known challenges even in the incompressible setting .
these challenges were bypassed in recently - developed temporal integrators for the incompressible fluctuating navier - stokes equations @xcite by avoiding the splitting inherent in projection methods .
extending this type of stokes - system approach to the low mach equations will be the subject of future research .
one of the principal motivations for developing the low mach number equations and our numerical implementation was to model recent experiments on the development of giant concentration fluctuations in the presence of sharp concentration gradients .
we first studied giant fluctuations in a time - independent or static setting , as observed experimentally by inducing a constant concentration gradient via a constant applied temperature gradient .
our simulations show that under conditions employed in experimental studies of the diffusive mixing of water and glycerol , it is reasonable to employ the boussinesq approximation . the results also indicate that the constant - transport - coefficient approximation that is commonly used in theoretical calculations is appropriate if the diffusion coefficient follows a stokes - einstein relation , but should be used with caution in general .
we continued our study of giant concentration fluctuations by simulating the temporal evolution of a rough diffusive interface during the diffusive mixing of hard disk fluids .
comparison between computationally - intensive event - driven molecular dynamics simulations and our hydrodynamic calculations demonstrated that the low mach number equations of fluctuating hydrodynamics provide an accurate coarse - grained model of fluid mixing .
special care must be exercised , however , in choosing the bare transport coefficients , especially the concentration diffusion coefficient , as these are renormalized by the fluctuations and can be strongly grid - dependent @xcite .
some questions remain about how to define and measure the bare transport coefficients from microscopic simulations , but we show that simply comparing particle and hydrodynamic calculations at large scales is a robust technique .
the strong coupling between velocity fluctuations and diffusive transport means that deterministic models have limited utility at mesoscopic scales , and even macroscopic scales in two - dimensions .
this implies that standard fluorescent techniques for measuring diffusion coefficients , such as fluorescence correlation spectroscopy ( fcs ) and fluorescence recovery after photobleaching ( frap ) @xcite , may not in fact be measuring material constants but rather geometry - dependent values @xcite .
fluctuating hydrodynamic simulations of typical experimental simulations , however , are still out of reach due to the very large separation of time scales between mass and momentum diffusion . surpassing this limitation requires the development of a semi - implicit temporal discretization that is stable for large time steps .
furthermore , it is also necessary to develop novel mathematical models and algorithms that are not only stable but also accurate in the presence of such large separation of scales .
this is a nontrivial challenge if thermal fluctuations are to be included consistently , and will be the subject of future research
. we would like to thank boyce griffith and mingchao cai for helpful comments .
j. bell , a. nonaka and a. garcia were supported by the doe applied mathematics program of the doe office of advanced scientific computing research under the u.s .
department of energy under contract no .
de - ac02 - 05ch11231 .
a. donev was supported in part by the national science foundation under grant dms-1115341 and the office of science of the u.s .
department of energy through early career award number de - sc0008271 .
t. fai wishes to acknowledge the support of the doe computational science graduate fellowship , under grant number de - fg02 - 97er25308 . y.
sun was supported by the national science foundation under award oci 1047734 .
as discussed in more depth in ref .
@xcite , there are fundamental mathematical difficulties with the interpretation of the nonlinear equations of fluctuating hydrodynamics due to the roughness of the fluctuating fields .
it should be remembered , however , that these equations are coarse - grained models with the coarse - graining length scale set by the size of the hydrodynamic cells used in discretizing the equations @xcite .
the spatial discretization removes the small length scales from the stochastic forcing and regularizes the equations .
it is important to point out , however , that imposing such a small - scale regularization _ _ ( smoothing ) of the stochastic forcing also requires a suitable renormalization of the transport coefficients @xcite , as we discuss in more detail in section [ sec : mixingmd ] .
as long as there are sufficiently many molecules per hydrodynamic cell the fluctuations in the spatially - discrete hydrodynamic variables will be small and the behavior of the nonlinear equations will closely follow that of the _ linearized _ equations of fluctuating hydrodynamics @xcite , which can be given a precise meaning @xcite .
it is therefore crucial to understand the linearized equations from a theoretical perspective , and to analyze the behavior of the numerical schemes in the linearized setting @xcite .
some of the most important quantities predicted by the fluctuating hydrodynamics equations are the equilibrium structure factors ( static covariances ) of the fluctuating fields .
these can be obtained by linearizing the compressible equations ( [ llns_primitive ] ) around a uniform reference state , @xmath425 , @xmath426 , @xmath427 , @xmath428 where @xmath429,\ ] ] and then applying a spatial fourier transform @xcite .
owing to fluctuation - dissipation balance the static structure factors are independent of the wavevector @xmath430 at thermodynamic equilibrium , @xmath431 note that density fluctuations do not vanish even in the incompressible limit @xmath25 unless @xmath43 .
while fluctuations in @xmath36 and @xmath196 are uncorrelated , the fluctuations in concentration and density are _ correlated _ even at equilibrium , @xmath432 we will see below that the low mach equations correctly reproduce the static covariances of density and concentration in the limit @xmath25 . the dynamics of the equilibrium fluctuations can also be studied by applying a fourier - laplace transform in time in order to obtain the dynamic structure factors ( equilibrium correlation functions ) as a function of wavenumber @xmath430 and wavefrequency @xmath52 @xcite .
it is well - known that the dynamic spectrum of density fluctuations @xmath433 exhibits three peaks for a given @xmath430 , one central rayleigh peak at small frequencies ( slow concentration fluctuations ) , and two symmetric brillouin peaks centered around @xmath54 .
as the fluid becomes less compressible ( i.e. , the speed of sound increases ) , there is an increasing separation of time - scales between the side and central spectral peaks .
as we will see below , the low mach equations reproduce the central peaks in the dynamic structure factors only , eliminating the side peaks and the associated stiff dynamics .
we now examine the spatio - temporal correlations of the steady - state fluctuations in the low mach number equations ( [ eq : momentum_eq],[eq : rho1_eq],[eq : div_v_constraint],[eq : rho_eq ] ) . in order to model the nonequilibrium setting in which giant concentration fluctuations
are observed , we include a constant background concentration gradient in the equations .
note that a density gradient will accompany a concentration gradient , and this can introduce some additional terms in @xmath434 depending on how @xmath435 depends on concentration . for simplicity , we assume @xmath435 is a constant so that the diffusive term @xmath436 in ( [ eq : rho1_eq ] ) is simply @xmath437 .
we also assume the viscosity @xmath10 is spatially constant , to get the simplified coupled velocity - concentration equations , @xmath438 where @xmath439 and @xmath124 is given by ( [ eq : eos_quasi_incomp ] ) .
we linearize the equations ( [ eq : simpl_primitive_eqs ] ) around a steady state , @xmath440 , @xmath441 , and @xmath442 , where the reference state is in mechanical equilibrium , @xmath443 .
we denote the background concentration gradient with @xmath444 .
we additionally assume that the reference state varies very weakly on length scales of order of the wavelength , an in particular , that @xmath70 and @xmath69 are essentially constant .
this allows us to drop the bars from the notation and employ a _ quasi - periodic _ or weak - gradient approximation @xcite . in the linear approximation
, the eos constraint relates density and concentration fluctuations , @xmath445 .
the term @xmath446 is second order in the fluctuations and drops out , but the advective term @xmath447 leads to a term @xmath448 in the concentration equation .
the forcing term due to gravity becomes @xmath449 . after a spatial fourier transform ,
the linearized form of ( [ eq : simpl_primitive_eqs ] ) becomes a collection of stochastic differential equations , one system of linear additive - noise equations per wavenumber , @xmath450.\label{eq : div_v_linearized}\end{aligned}\ ] ] replacing the right hand side of ( [ eq : div_v_linearized ] ) with zero leads to the incompressible approximation used in ref .
@xcite , corresponding to the boussinesq approximation of taking the limit @xmath277 while keeping the product @xmath276 constant .
let us first compare the dynamics of the equilibrium fluctuations ( @xmath451 ) in the low mach equations with those in the complete compressible equations .
for simplicity of notation we will continue to use the hat symbol to denote the space - time fourier transform . in the wavenumber - frequency @xmath452 fourier domain , the concentration fluctuations in the absence of a gradient are obtained from ( [ eq : c_t_linearized ] ) , @xmath453 which is the same as the compressible equations .
the density fluctuations follow the concentration fluctuations , @xmath454 , and the dynamic structure factor for density shows the same central rayleigh peak as obtained from the isothermal compressible equations @xcite , @xmath455 where we used eq .
( [ stoch_flux_covariance ] ) for the covariance of @xmath456 .
this shows that the low mach number equations correctly reproduce the slow fluctuations ( small @xmath52 ) in density and concentration , while eliminating the side brillouin peaks associated with the fast isentropic pressure fluctuations . the fluctuations in velocity , however , are different between the compressible and low mach number equations .
let us first examine the transverse ( solenoidal ) component of velocity @xmath457 , where @xmath458 is the constant - density orthogonal projection onto the space of divergence - free velocity fields ( @xmath459 in fourier space ) .
applying the projection operator to the velocity equation ( [ eq : v_t_linearized ] ) shows that the fluctuations of the solenoidal modes are the same as in the incompressible approximation , @xmath460 the fluctuations of the compressive velocity component @xmath461 , on the other hand , are driven by the stochastic mass flux @xmath456 , as seen from eq .
( [ eq : div_v_linearized ] ) at thermodynamic equilibrium , @xmath462 the dynamic structure factor ( space - time fourier spectrum ) of the longitudinal component @xmath463 does not decay to zero as @xmath53 .
this indicates that the fluctuations of velocity are not only white in space but also white in time . in the incompressible approximation
@xmath277 so that the longitudinal velocity fluctuations vanish and the static spectrum of the velocity fluctuations is equal to the projection operator , @xmath464 @xcite . in the compressible equations ,
the dynamic structure factor for the longitudinal component of velocity decays to zero as @xmath53 because it has two sound ( brillouin ) peaks centered around @xmath465 , in addition to the central diffusive ( rayleigh ) peak .
the low mach number equations reproduce the central peak ( slow fluctuations ) correctly , replacing the side peaks with a flat spectrum for large @xmath52 .
the origin of this unphysical behavior is the unjustified assumption of infinite separation of time scales between the propagation of sound and the diffusion of mass , momentum and energy . in reality
, the same molecular motion underlies all of these processes and the incompressible or the low mach number equations can not be expected to reproduce the correct physical behavior at very short time scales ( @xmath466 ) .
if we neglect the term involving @xmath456 in ( [ eq : div_v_linearized ] ) and eliminate the lagrange multiplier ( non - thermodynamic pressure ) @xmath31 using ( [ eq : div_v_linearized ] ) , we obtain the linearized velocity equation in fourier space @xmath467\v k+i\beta\chi\left(\nu-\chi\right)k^{2}\left(\widehat{\delta c}\right)\v k.\label{eq : vhat_t_proj}\end{aligned}\ ] ] it is straightforward to obtain the steady - state covariances ( static structure factors ) in the presence of a concentration gradient from the linearized system of velocity - concentration equations ( [ eq : c_t_linearized],[eq : vhat_t_proj ] ) @xcite
. the procedure amounts to solving a linear system for three covariances ( velocity - velocity , concentration - concentration , and velocity - concentration ) .
these types of calculations are particularly well - suited for modern computer algebra systems like maple and can be carried out for arbitrary wavenumber and background concentration gradient .
we omit the full solution for brevity .
experiments measure the steady - state spectrum of concentration fluctuations averaged along the gradient @xcite , and we will therefore focus on wavenumbers perpendicular to the gradient , @xmath468 .
a straightforward calculation shows that the concentration fluctuations are enhanced as the square of the applied gradient , @xmath469}\ , h_{\parallel}^{2},\label{eq : s_c_c}\ ] ] where * @xmath470 * and @xmath471 denote the perpendicular and parallel component relative to gravity , respectively .
the term in the denominator involving @xmath472 comes from the low mach number constraint ( [ eq : div_v_constraint ] ) and is usually negligible since the concentration gradient is parallel to gravity or @xmath473 . without this term
the result ( [ eq : s_c_c ] ) is the same result as obtained in @xcite , and shows that fluctuations at wavenumbers below @xmath474 are suppressed by gravity , as we study numerically in section [ sec : giantfluct ] .
in our spatial discretization , we use centered differencing for the advective terms because this leads to a skew - adjoint discretization of advection @xcite that maintains discrete fluctuation - dissipation balance in the spatially - discretized stochastic equations @xcite .
it is well - known that centered discretizations of advection do not preserve monotonicity properties of the underlying pdes in the deterministic setting , unlike one - sided ( upwind ) discretizations .
therefore , our spatio - temporal discretization can lead to unphysical oscillations of the concentration and density in cases where the cell peclet number @xmath475 is large . in the deterministic setting ,
@xmath476 can always be decreased by reducing @xmath128 and resolving the fine scale dissipative features of the flow . however , in the stochastic setting , the magnitude of the fluctuating velocities at equilibrium is @xmath477 where @xmath156 is the volume of the hydrodynamic cell .
therefore , in two dimensions the characteristic advection velocity magnitude is @xmath478 .
this means that in two dimensions @xmath476 is independent of the grid size and reducing @xmath128 can not fix problems that may arise due to a large cell peclet number . for some of the simulations reported in section [ sec : mixingmd ] ,
we have found it necessary to implement a spatial filtering procedure to reduce the magnitude of the fluctuating velocities while preserving their spectrum as well as possible at small wavenumbers .
the filtering procedure consists of applying a local averaging operation to the spatially - discretized random fields @xmath479 and @xmath24 independently along each cartesian direction .
this local averaging smooths the random forcing and thus reduces the spectrum of the random forcing at larger wavenumbers .
the specific filters we use are taken from ref .
for stencil width @xmath480 , filtering a discrete field @xmath481 in one dimension takes the form @xmath482 in fourier space , for discrete wavenumber @xmath483 this local averaging multiplies the spectrum of @xmath481 by @xmath484 and therefore maintains the second - order accuracy of the spatial discretization . at the same time , the filtering reduces the variance of the fluctuating fields by about a factor of two in one dimension ( a larger factor in two dimensions ) .
the spectrum of the fluctuations can be preserved even more accurately if a stencil of width @xmath380 is used for the local averaging , @xmath485 giving a sixth - order accurate filter @xmath486 and a reduction of the variance by about a third in one dimension . in two and three dimensions the filtering
operators are simple tensor products of one - dimensional filtering operators .
note that we only use these filters with periodic boundary conditions .
one can , of course , also use fourier transform techniques to filter out high frequency components from the stochastic mass and momentum fluxes .
the hydrodynamic simulations described in section [ sec : mixingmd ] require as input transport coefficients , notably , the shear viscosity @xmath10 and diffusion coefficient @xmath12 , which need to be extracted from the underlying microscopic ( molecular ) dynamics .
this is a very delicate and important step that has not , to our knowledge , been carefully performed in previous studies . in this appendix
we give details about the procedure we developed for this purpose . as discussed in more detail in refs .
@xcite , the transport coefficients in fluctuating hydrodynamics are not universal material constants but rather depend on the spatial scale ( degree of coarse - graining ) under question .
we emphasize that this scale - dependent renormalization is not a molecular scale effect but rather an effect arising out of hydrodynamic fluctuations , and persists even at the hydrodynamic scales we are examining here .
the best way to define and measure transport coefficients is by examining the dynamics of _ equilibrium _ fluctuations , specifically , by examining the _ dynamic structure factors _ of the hydrodynamic fields @xcite , i.e. , the equilibrium averages of the spatio - temporal fourier spectra of the fluctuating hydrodynamic fields . for a hydrodynamic variable @xmath488 that is transported by a purely diffusive process
, the spectrum of the fluctuations at a given wavenumber @xmath191 and wavefrequency @xmath52 is expected to be a lorentzian peak of the form @xmath489^{-1},\ ] ] where in general the diffusion constant @xmath490 depends on the the wavenumber @xmath191 ( wavelength @xmath491 ) .
we can therefore estimate the diffusion coefficient @xmath12 by fitting a lorentzian peak to @xmath492 for different @xmath191 s ( i.e. , @xmath493 ) .
similarly , we can estimate the kinematic viscosity @xmath439 by fitting a lorentzian curve to dynamic structure factors for the scaled vorticity , @xmath494 .
we performed long equilibrium molecular dynamics simulations of systems corresponding to a grid of @xmath495 hydrodynamic cells , and then calculated the discrete spatio - temporal fourier spectrum of the hydrodynamic fields at a collection of discrete wavenumbers @xmath430 .
since these simulations are at equilibrium , the systems are well - mixed , specifically , the initial configurations were generated by randomly assigning a species label to each particle .
we then performed a nonlinear least squares lorentzian fit in @xmath52 for each @xmath430 and estimated the width of the lorentzian peak .
the results for the dynamics of the equilibrium vorticity fluctuations are shown in fig .
[ fig : s_kw_hdmd ] .
we see that kinematic viscosity is relatively constant for a broad range of wavelengths , consistent with fluctuating hydrodynamics calculations @xcite and previous molecular dynamics simulations @xcite . for the pure component one fluid , @xmath60 , with density
@xmath496 the figure shows @xmath497 .
we therefore used @xmath498 in all of the hydrodynamic runs reported in section [ sec : mixingmd ] .
this is about @xmath499 higher than the prediction of the simple enskog kinetic theory @xcite , @xmath500 , and is consistent with the estimates reported in ref .
because of the diffusion coefficient is small at the densities we study , more specifically , because the schmidt number @xmath501 is larger than 10 , we were unable to obtain reliable estimates for @xmath502 from the dynamic structure factor for concentration .
simple dimensional analysis or kinetic theory shows that @xmath503 .
since the disks of the two species have equal diameters the viscosity of the pure second fluid component is @xmath504 there is no simple theory that accurately predicts the concentration dependence of the viscosity of a hard disk mixture at higher densities @xcite . to our knowledge
there is no published enskog kinetic theory calculations for hard - disk mixtures in two dimensions , even for the simpler case of equal diameters . as an approximation to the true dependence , we employed a simple linear interpolation of the _ kinematic _ viscosity @xmath505 as a function of the mass concentration @xmath34 between the two known values @xmath506 and @xmath507 . the numerical results for mixtures with mass ratios @xmath375 and @xmath378 in fig .
[ fig : s_kw_hdmd ] are consistent with this approximation to within the large error bars .
for example , for @xmath508 and @xmath378 the interpolation gives @xmath509 which is in reasonable agreement with the numerical estimate .
estimates of the momentum diffusion coefficient ( viscosity ) @xmath439 obtained from the width of the central peak in the dynamic structure factor of vorticity . a collection of @xmath510 distinct discrete wavenumbers @xmath430 were used and the width of the peaks estimated using a nonlinear least squares lorentzian fit.,scaledwidth=75.0% ] for the inter - species diffusion coefficient @xmath12 , which we emphasize is distinct from the self - diffusion coefficients for particles of either species , enskog kinetic theory predicts no concentration dependence and a simple scaling with the mass ratio @xcite , @xmath511 this particular dependence on mass ratio @xmath363 comes from the fact that the average relative speed between particles of different species is @xmath512 , where @xmath513 is the reduced molecular mass .
we have assumed in our hydrodynamic calculations that the diffusion coefficient is independent of concentration and follows ( [ eq : chi_r_scaling ] ) .
the only input to the hydrodynamic calculation is the bare self - diffusion coefficient for the pure component fluid , @xmath514 .
diffusion is strongly renormalized by thermal fluctuations , and fluctuating hydrodynamics theory and simulations predict a strong dependence of the diffusion coefficient @xmath12 on the wavelength @xcite , consistent with molecular dynamics results @xcite . in order to estimate the appropriate value of the bare diffusion coefficient @xmath299 we numerically solved an inverse problem . using simple bisection
, we looked for the value of @xmath299 that leads to best agreement for the average or `` macroscopic '' diffusion ( mixing ) between the particle and continuum simulations .
specifically , we calculated the density of the first species @xmath391 along the @xmath203-direction by averaging @xmath36 in each horizontal row of hydrodynamic cells , see eq .
( [ eq : rho_1_h ] ) .
the results for @xmath515 for mass ratios @xmath345 and @xmath378 are shown in fig . [
fig : spreading_64x64 ] at different points in time for systems of size @xmath358 cells .
the figures show the expected sort of diffusive mixing profile , and is exactly what would be used in experiments to measure diffusion coefficients using fluorescent techniques such as fluorescence recovery after photo - bleaching ( frap ) @xcite .
this macroscopic measurement smooths over the fluctuations ( roughness ) of the diffusive interface and only measures an effective diffusion coefficient at the scale of the domain length @xmath336 .
if deterministic hydrodynamics is employed , @xmath391 is the solution of a one - dimensional system of equations obtained by simply deleting the stochastic forcing and the @xmath201-dependence in the low mach equations . instead of solving this system analytically , we employed our spatio - temporal discretization with fluctuations turned off , and with an effective diffusion coefficient @xmath516 that accounts for the renormalization of the diffusion coefficient by the thermal fluctuations .
( left panel ) diffusive evolution of the horizontally - averaged density @xmath391 for a system of size @xmath358 hydrodynamic cells and density ratio @xmath345 , as obtained from hdmd simulations ( circles , averaged over 64 runs ) , deterministic hydrodynamics with @xmath517 ( dashed lines ) , and fluctuating hydrodynamics with @xmath518 ( squares , averaged over 64 runs ) .
error bars are comparable to the symbol size and not shown for clarity .
( right panel ) same as the left panel except the density ratio is @xmath375 and the transport coefficients are adjusted according to ( [ eq : eta_r_scaling],[eq : chi_r_scaling]).,title="fig:",scaledwidth=49.0%](left panel ) diffusive evolution of the horizontally - averaged density @xmath391 for a system of size @xmath358 hydrodynamic cells and density ratio @xmath345 , as obtained from hdmd simulations ( circles , averaged over 64 runs ) , deterministic hydrodynamics with @xmath517 ( dashed lines ) , and fluctuating hydrodynamics with @xmath518 ( squares , averaged over 64 runs ) .
error bars are comparable to the symbol size and not shown for clarity .
( right panel ) same as the left panel except the density ratio is @xmath375 and the transport coefficients are adjusted according to ( [ eq : eta_r_scaling],[eq : chi_r_scaling]).,title="fig:",scaledwidth=49.0% ] by matching the profile @xmath391 between the hdmd and the fluctuating and deterministic hydrodynamic simulations at mass ratio @xmath345 and system size @xmath358 cells , we obtained estimates for the bare @xmath299 and the renormalized @xmath390 coefficient ( see fig .
[ fig : spreading_64x64 ] ) .
the best estimate for the bare diffusion coefficient based on this matching in the absence of filtering is @xmath519 .
this compares reasonably - well to the prediction of enskog theory @xcite of @xmath520 , as well as to the measurement of the self - diffusion coefficient for a periodic system with 169 disks reported in ref .
@xcite , @xmath521 ( recall that a single hydrodynamic cell in our case contains about 76 particles ) . when a 5-point filter is employed
the estimate is @xmath522 and when a 9-point filter is employed @xmath523 .
the estimated renormalized diffusion coefficient is much larger , @xmath524 , consistent with a rough estimate based on the simple theory presented in ref .
@xcite , @xmath525 to within statistical accuracy we were not able to detect the increase in the estimated diffusion coefficients when using the larger systems of size @xmath359 cells , however , for @xmath495 it was clear that @xmath390 is reduced .
it is important to emphasize that @xmath390 is not a material constant but rather depends on the details of the problem in question , in particular , the system geometry and size and boundary conditions @xcite .
by contrast , @xmath299 is a constant for a given spatial discretization , and one can use the same number for different scenarios so long as the hydrodynamic cell size and the filter are kept fixed . unlike deterministic hydrodynamics , which presents an incomplete picture of diffusion , fluctuating hydrodynamics correctly accounts for the important contribution of the thermal velocity fluctuations and the roughness of the diffusive interface seen in fig . [
fig : mixingillustrationmass4 ] .
a. naji , p. j. atzberger , and frank l. h. brown .
hybrid elastic and discrete - particle approach to biomembrane dynamics with application to the mobility of curved integral membrane proteins .
, 102(13):138102 , 2009 .
y. hennequin , d. g. a. l. aarts , j. h. van der wiel , g. wegdam , j. eggers , h. n. w. lekkerkerker , and d. bonn .
drop formation by thermal fluctuations at an ultralow surface tension .
, 97(24):244502 , 2006 . | we formulate low mach number fluctuating hydrodynamic equations appropriate for modeling diffusive mixing in isothermal mixtures of fluids with different density and transport coefficients .
these equations represent a coarse - graining of the microscopic dynamics of the fluid molecules in both space and time , and eliminate the fluctuations in pressure associated with the propagation of sound waves by replacing the equation of state with a local thermodynamic constraint .
we demonstrate that the low mach number model preserves the spatio - temporal spectrum of the slower diffusive fluctuations .
we develop a strictly conservative finite - volume spatial discretization of the low mach number fluctuating equations in both two and three dimensions and construct several explicit runge - kutta temporal integrators that strictly maintain the equation of state constraint .
the resulting spatio - temporal discretization is second - order accurate deterministically and maintains fluctuation - dissipation balance in the linearized stochastic equations .
we apply our algorithms to model the development of giant concentration fluctuations in the presence of concentration gradients , and investigate the validity of common simplifications such as neglecting the spatial non - homogeneity of density and transport properties .
we perform simulations of diffusive mixing of two fluids of different densities in two dimensions and compare the results of low mach number continuum simulations to hard - disk molecular dynamics simulations .
excellent agreement is observed between the particle and continuum simulations of giant fluctuations during time - dependent diffusive mixing .
# 1 # 1 # 1 # 1#1 # 1#1 # 1#1 # 1|#1| # 1#1 # 1#1 |
the use of quantum master equations ( me ) for weakly coupled systems is well - spread , notably in various resonance spectroscopies and in experimental realization of ensemble quantum computing @xcite .
the traditional system plus bath models seek to derive an me to describe the non - unitary behavior of the system by tracing out the bath degrees of freedom @xcite . though the mes typically involve a second - order perturbative expansion of the system - bath coupling , the application of any external drive on the system is usually treated in the linear response regime @xcite .
a parallel approach assumes that the coupling and the drive induce independent rates of variation in the system resulting in a unitary evolution due to the drive @xcite .
therefore the system - bath coupling remains the only source of damping and hence non - unitary behavior in such systems . on the contrary , the semi - classical theory of bloembergen , purcell and pound hints at the existence of nonlinear response terms in the mes which lead to additional damping of the population differences @xcite .
other semi - classical models used to explain the phenomenon of saturation also indicate the presence of higher order terms of the drive @xcite .
but the existing prescriptions to derive mes are not well - suited to obtain time - nonlocal higher - order contributions of the drive . here
we attempt to resolve this conflict by considering the finite memory effects of the bath originating from the equilibrium fluctuations of its energy levels .
the bath - memory renders the higher order perturbations of the drive to be time - nonlocal and therefore opens up the possibility of a perturbative expansion up to any desired order .
restricting ourselves upto second - order , we subsequently derive a modified form of the bloch equations having drive - induced damping terms and also provide a direct experimental verification of its validity .
to begin with , we consider an ensemble , each member of which is described by the following hamiltonian in the angular frequency unit ( t ) = _ ^s +
_ ^l + ^sl + ^s(t ) + ^l(t ) , where , @xmath0 and @xmath1 represent the static hamiltonians of the relevant and the non - relevant degrees of freedom respectively , which are weakly coupled by the term @xmath2 . the collection of the relevant degrees of freedom constitute our system of interest whereas the non - relevant degrees of freedom together form a heat bath .
@xmath3 denotes the external drive to the system .
it is assumed that the bath is in thermal equilibrium at an inverse temperature @xmath4 and experiences equilibrium fluctuations @xmath5 , which solely act on the bath .
@xmath5 may assume different values for different ensemble members while respecting the constraints imposed by the requirement of sustained thermal equilibrium .
apart from @xmath5 , the other terms in the hamiltonian are identical for all ensemble members and as such @xmath1 determines the equilibrium density matrix of the bath .
since the fluctuations do not drive the bath away from equilibrium , we choose @xmath5 to be diagonal in the eigen - basis @xmath6 of @xmath1 , represented by @xmath7 .
@xmath8-s are assumed to be independent , gaussian , @xmath9-correlated stochastic variables with zero mean and standard deviation @xmath10 .
we further assume that the fluctuation @xmath5 and the state experiencing it at the time instant @xmath11 are statistically independent so as to preserve causality . starting with this description , we perform all subsequent calculations in the interaction representation of @xmath12 and all hamiltonians in this representation are denoted by @xmath13 with relevant superscripts .
dynamics of a chosen member of the ensemble , denoted by @xmath14 in this representation , is described by the liouville - von neumann equation : [ lv ] _
t((t)(t ) ) = - , where @xmath15 . to derive an me for the system we first take trace over the irrelevant degrees of freedom from both sides of the above equation and then perform a desired number of iterative substitutions using the integral form of the schrdinger equation , @xmath16 , @xmath17 .
( appendix : a ) .
the main motivation behind adopting this iterative procedure is to capture the effect of a finite memory in the irrelevant part of the state function due to the fluctuations .
we thus have [ m3 ] _
t_1(tr_l(t_1)(t_1 ) ) & = & -tr_l-_t^t_1 dttr_lh^eff(t_1)h^eff(t)(t)(t_1 ) + & & -h^eff(t)(t)(t_1)h^eff(t_1)-h^eff ( t_1)(t_1)(t)h^eff(t ) + & & + ( t_1)(t)h^eff(t)h^eff(t_1 ) -_t^t_1 dttr_l + & & + _
t^t_1 dt_t^t_1 dttr_l + & & + _
t^t_1 dt_t^t_1 dttr_l + [ ( h^eff)^3(t ) ] , where @xmath18 . averaging over the ensemble while using the statistical properties of @xmath19 , then results in a dynamical equation for the system , involving the effective hamiltonian @xmath20 and terms of the form @xmath21 in second and higher orders of @xmath20 , where , the angular brackets denote ensemble average ( appendix : b ) .
usually the largest time - scales associated with the correlation of the bath variables ( @xmath22 ) is much smaller than that of the system @xmath23 which is determined by @xmath20 . as such , there exists two widely different time - scales in this problem which allows a coarse - graining over the time interval @xmath24 $ ] , such that @xmath25 .
the second order terms in @xmath26 as well as @xmath27 are smaller than the first order terms as long as such a time - scale separation exists in the problem . third and
higher order terms are successively smaller by factors of @xmath28 or @xmath29 .
hence we neglect terms @xmath30 $ ] for a sufficiently weak drive and coupling @xcite . in order to evaluate @xmath21 @xmath31
, we use the time - ordered propagator generated only by @xmath19 , since in the time - scale of the course - grain interval , only the bath fluctuations are significant .
this , in essence is similar to the markov approximation , whereby one neglects the time evolution of the system in the second - order term of the me , within the short duration @xmath32 @xcite .
any back - action of the system dynamics to the bath via the coupling is avoided by this protocol . to obtain expressions in a familiar closed form
we ascribe an identical standard deviation @xmath33 to all @xmath8-s . assuming that at the beginning of the coarse - grain interval , the density matrix for the whole ensemble can be factorized into that of the system and the bath with the latter at thermal equilibrium , we obtain [ ddm1 ] ( t_1)(t_2)= _ s(t)_l^eq(-^2t_1 - t_2 ) , where @xmath34 denotes the density matrix of the system whereas @xmath35 denotes the equilibrium density matrix of the heat bath ( appendix :
c ) . using the above result
we find that the integrand in the second order term of the coarse - grained equation takes the form of a double commutator decaying within the time - scale of @xmath36 .
thus @xmath36 forms the upper bound of the time - scales during which the bath correlations are significant and as such we replace it by @xmath22 . following the prescription of cohen - tannoudji _
et.al._ , we then arrive at the following me : [ mf ] _
t_s(t ) & = & -^ -_0^ d ^ e^-_c , where , the superscript ` sec ' denotes that only the secular contributions are retained ( ensured by the coarse graining ) and @xmath37 denotes the partial trace . unlike the usual forms of me found in literature , equation ( [ mf ] )
has a finite , time - nonlocal , second - order contribution of the external drive to the system @xcite .
the derived me ( [ mf ] ) yields lorenztian spectral density functions due to the presence of the exponential decay term . while the equation ensures that the system approaches equilibrium , the specific form of the equilibrium distribution is determined by the resonance condition ( appendix : d ) . as a practical example
, we now consider our system to be the spin degrees of freedom of an ensemble of isolated spin @xmath38 nuclei ( qubits ) embedded in solute molecules of a dilute liquid solution placed in a static external magnetic field , the direction of which is chosen to define our @xmath39 axis .
the spatial degrees of freedom relevant in this problem constitute an isotropic thermal bath at an inverse temperature @xmath4 .
the system is also subjected to a weak , near - resonant magnetic field ( the drive ) of strength @xmath40 along @xmath41 , oscillating with a frequency @xmath42 . in a typical nmr measurement , heterodyne detection
is performed followed by low - pass filtering , which is equivalent to measurements made in a co - rotating frame of frequency @xmath42 @xcite .
hence we suitably define the macroscopic magnetization components @xmath43 @xmath44 . using the me ( [ mf ] ) along with the assumption of isotropicity _
i.e. _ @xmath45 = 0 $ ] , we arrive at a modified form of the bloch equations for @xmath43 : [ mbl ] _ z & = & _ 1m_y -(m_z - m _ ) - m_z , + + _ x & = & m_y - m_x , + + _ y & = & -m_x - _ 1m_z -m_y - m_y + where , @xmath46 is the offset from the resonance frequency @xmath47 , @xmath48 are the longitudinal and transverse relaxation times respectively and @xmath49 is the equilibrium magnetization @xcite . in deriving the above equations we have neglected the counter - rotating components of the driving field and also all lamb shift terms @xcite ( appendix : e ) .
the additional terms in the equations ( [ mbl ] ) proportional to @xmath50 lead to a faster approach to saturation in conformity with @xcite . if @xmath51 , _
i.e. _ at high motional - narrowing limit , these extra terms become negligible and thereby we recover the original bloch equations .
is equivalent to an on - resonance @xmath52-pulse block composed of @xmath53 pulses along @xmath54 respectively with @xmath55 .
@xmath56 in the figure denotes the drive applied along @xmath57 .
the 3d diagrams atop the pulses show how the refocusing happens irrespective of the actual value of the flip angle @xmath58 and hence @xmath40 .
the pulse sequence below shows the operations performed on the proton ( @xmath59h ) and the gradient channels respectively . after a recycle delay of @xmath60 , we employ a super - cycle of r@xmath61 inspired by waltz-8 as the drive sequence @xcite .
@xmath62 indicates all pulses in r@xmath61 are applied along inverted axis .
@xmath63^n$ ] implies that the sequence within the brackets is repeated @xmath64 times consecutively and @xmath65 denotes the length of a super - cycle . for detection
, we have applied a gaussian selective pulse of flip angle @xmath66 in the presence of the applied gradient @xmath67 to facilitate slice selection using spatial encoding . immediately after the slice selection
, an opposite compensatory gradient @xmath68 is used to minimize phase distortion during slice - selection @xcite . the free induction decay ( fid )
has been recorded in the absence of any gradient .
magnetization , @xmath69 is measured from the fid as a function of @xmath64 . from the @xmath69 versus @xmath70 data we measure
@xmath71 using least - square curve fit . ]
the resonance condition @xmath72 decouples the dynamics of @xmath73 from that of @xmath74 and @xmath69 . when @xmath75 , we get a damped nutation in the @xmath76-plane , starting from @xmath39 .
the nutation frequency is @xmath77 and the damping rate is @xmath78 . since @xmath22 is of the order of ` ps ' in liquids and typically @xmath79khz ,
the term @xmath50 is usually small in comparison to @xmath80 @xcite .
thus in order to observe its effect , we drive the system for long enough times ( of the order of @xmath81ms ) .
the drive strength , @xmath40 is not entirely homogeneous throughout the sample volume in practical cases .
this leads to an additional decay of the measured signal due to the fanning out of the magnetization components in the @xmath76-plane .
usually the measured decay of nutation is dominated by this first - order dephasing , due to the presence of drive - inhomogeneity in the sample , which obscures the @xmath82 dependence of the damping rate . to avoid this
we adopt a refocusing scheme so that @xmath83 at every instant of measurement @xmath11 . under these assumptions
we have a solution of the form @xmath84 $ ] and @xmath85 , where @xmath86 , @xmath87 and @xmath88 are functions of @xmath49 , @xmath89 , @xmath90 , @xmath40 and @xmath22 .
so we measure the decay rates , @xmath71 at various drive strengths , @xmath40 to establish the quadratic dependence of the former on the latter .
an efficient refocusing process should aim to make @xmath91 as small as possible at each measuring instant @xmath11 .
the simplest possible method to accomplish this involves the application of a continuous train of pulses with flip - angles @xmath58 and @xmath92 ( i.e. @xmath93 ) alternately with measurements made after even number of pulses .
nutation by an angle @xmath92 is achieved by applying the drive along @xmath94 . since
after an even number of pulses in the sequence , the net phase of the nutating magnetization components become zero , the resulting dynamics remains immune to the drive inhomogeneity - induced dephasing .
we choose @xmath55 for our experiments to minimize precession about @xmath39 ( if any , due to incorrect shim profile ) which takes the magnetization out of the @xmath76-plane ( during the finite rise and fall times of the pulses ) .
the @xmath52-pulse block , @xmath95 is more efficient than the simple @xmath96 block in minimizing the @xmath97-leakage , @xmath91 .
this can be verified by calculating the ratio of the leakage magnetizations , @xmath98 obtained by running the two sequences for the same duration , @xmath99 ( i.e. @xmath100 is run twice consecutively while @xmath101 is run only once ) .
this ratio turns out to be @xmath102 $ ] .
now , @xmath103 and @xmath104 as mentioned before and since for small @xmath58 , @xmath105\sim\theta$ ] , the ratio @xmath98 is negligible .
inspired by the waltz-@xmath106 decoupling scheme , we find , through simulations with @xmath107s , that the super - cycle @xmath108 , ( where @xmath109 ) is more effective than the simple @xmath52-pulse block @xmath101 in minimizing @xmath91 and thus we use it as our driving protocol @xcite . in order to further eliminate the effect of drive - inhomogeneity , we select a thin slice ( @xmath110 mm ) near the middle of the sample ( effective height @xmath111 mm ) , for detection , by using a selective gaussian @xmath112 pulse along @xmath41 in presence of an applied @xmath113-gradient of about @xmath114 @xcite .
the duration of the positive gradient is @xmath115ms whereas that of the compensatory gradient is @xmath116ms .
the experimental protocol is illustrated in fig .
[ refocussing - scheme ] . to experimentally verify the equations ( [ mbl ] )
, we choose the singlet imine protons in a dilute ( mm ) solution of n-(4@xmath117-methoxybenzylidene)-4-n - butylaniline in dmso - d@xmath118 placed in a static magnetic field along @xmath39 , as our target spin system .
the application of an on - resonance ( @xmath119 ) drive along @xmath41 , results in a damped nutation in the @xmath76-plane ( if @xmath75 ) , starting from @xmath39 , with a damping rate : @xmath120 . versus the drive strength @xmath40 .
the upright triangles denote the experimentally determined values of @xmath71 with the vertical bars denoting the errors in the determination of @xmath71 at 95@xmath121 level of confidence .
the solid line ( color online ) shows the parabolic fit of the experimental data . ]
we have performed all experiments on a bruker avance iii @xmath122 t nmr spectrometer at @xmath123k . the drive strength @xmath124 has been varied from @xmath125khz to @xmath126khz , in steps of @xmath127khz . for each drive strength
, we have determined the decay rate of nutation , @xmath71 as a function of nutation period . a plot of @xmath71 as a function of @xmath40
is shown in fig .
the chemical shift of the chosen imine protons is @xmath128ppm w.r.t .
tms @xcite .
the choice of the molecule is motivated by its relatively slow isotropic rotation in solution phase due to its long structure .
the @xmath89 and @xmath90 relaxation times for our system , measured using standard techniques , are @xmath129s and @xmath130s respectively @xcite . for each value of @xmath131
, we drive the system @xmath64 times , where @xmath64 ranges from @xmath132 to @xmath133 in steps of @xmath134 , so that the maximum drive - time does not exceed @xmath135ms .
thus for @xmath136khz the maximum value of @xmath64 is @xmath137 . for all other values of @xmath138
we use the full range of @xmath64 .
this ensures that imperfections due to finite rise and fall times of the pulses have nearly identical effects for each @xmath138 .
while @xmath139khz is approximately close to the maximum power limit of the spectrometer , we do not go below the 3khz limit simply because the drive time exceeds @xmath135ms for a significantly small value of @xmath64 compared to @xmath133 thereby leading to distortions .
we calculate @xmath69 after each experimental run from the fid using plancherel s theorem @xcite . taking the natural logarithm of the time - series of @xmath69 and fitting the result with a straight line
we obtain @xmath71 for a particular value of @xmath40 .
we note that our experimental protocol is sensitive to shim conditions and the accurate calibration of the drive power . in all the experiments we calibrate our pulses using the fact that a @xmath140db pulse produces a nutation frequency of @xmath141khz .
the success of the refocusing scheme also relies on minimal translational diffusion during @xmath142 which we presume to be the case since the duration of @xmath142 is of the order of tens of @xmath143s .
the maximum drive strength used in our experiments is @xmath144khz ( close to the maximum permissible power ) whereas the proton larmor frequency of the spectrometer is of about @xmath135mhz .
thus the second order approximation in our me is valid for all practical considerations .
the @xmath71 versus @xmath40 data is fit with a parabola of the form @xmath145 as shown in fig .
2 . from the fit
we obtain the value of @xmath22 to be @xmath146s .
the value of @xmath147 is found to be @xmath148hz which is same as the value of @xmath80 within experimental error , the latter being estimated using standard measurement techniques @xcite . for the @xmath71 versus @xmath82 behavior , we find fair agreement between the theoretical prediction and the experimental observation .
hence we infer that our experimental results confirm the validity of the dynamical equations ( [ mbl ] ) obtained from our me ( [ mf ] ) .
our results reveal that the finite memory of the bath plays a central role in giving rise to an additional damping of rabi oscillations in the form of a time - nonlocal nonlinear response of the system to the drive .
application of our formalism to magnetic resonance results in a modified form of bloch equations having drive - induced damping terms .
since @xmath22 is a measure of the bath correlation times , such effects may become significant for systems coupled to thermal baths with large @xmath22 ; the magnetic resonance of solid samples is a case in point . for liquids ,
such nonlinear effects are small and is usually obscured by the drive inhomogeneity . however a refocusing scheme which prevents the dephasing due to such inhomogeneities during nutation , helps confirm the existence of the nonlinear terms . at extreme motional - narrowing limit ( @xmath149 ) ,
the equations ( [ mbl ] ) predict an expected unitary nutation of the system due to the drive .
our choice of @xmath9-correlated fluctuations lead to an exponential memory function ; although other noise models may also serve the purpose .
also , our assumption of identical standard deviations @xmath10 of the bath energy levels is not rigorous but it may be reasonable for the levels which are effectively coupled to the system . since @xmath150 , this damping becomes small for smaller @xmath22 and hence it is envisaged that a bath with stronger fluctuations ( towards the extreme motional narrowing limit ) would make the system more immune to drive - induced damping .
damping of this kind may pose a serious hindrance in the realization of long - time quantum information processing protocols in the context of ensemble computing .
dynamics of a chosen member of the ensemble , denoted by @xmath14 in the interaction representation is described by the liouville - von neumann equation : [ alv ] _
t((t)(t ) ) = - , where @xmath15 . to describe the system dynamics ,
we take a partial trace over the irrelevant degrees of freedom on both sides of equation ( [ alv ] ) .
since @xmath19 is an operator on the irrelevant subspace , the partial trace removes its commutator in the r.h.s . of the equation ( 2 ) which then becomes @xmath151 $ ] , where @xmath152 .
@xmath153 denotes the partial trace .
using the fact that @xmath154 for @xmath155 we obtain [ am ] _
t_1(tr_l(t_1)(t_1 ) ) & = & -tr_l .
further expansion of the first - order terms in equation ( [ am ] ) leads to the exact expression [ am1 ] _
t_1(tr_l(t_1)(t_1 ) ) & = & -tr_l + & -&_t^t_1 dt tr_lh^eff(t_1)h^all(t)(t)(t)-h^all(t)(t)(t)h^eff(t_1 ) + & -&h^eff(t_1)(t_1)(t)h^all(t ) + ( t_1)(t)h^all(t)h^eff(t_1 ) . substituting @xmath156 in the integrand of equation ( [ am1 ] ) by @xmath157 , using the integral form of the schrdinger equation results in [ am2 ] _
t_1(tr_l(t_1)(t_1 ) ) & = & -tr_l-_t^t_1 dttr_lh^eff(t_1)h^eff(t)(t)(t_1 ) + & & -h^eff(t)(t)(t_1)h^eff(t_1)-h^eff ( t_1)(t_1)(t)h^eff(t ) + & & + ( t_1)(t)h^eff(t)h^eff(t_1 ) -_t^t_1 dttr_lh^eff(t_1)h^l(t)(t)(t_1 ) + & & -h^l(t)(t)(t_1)h^eff(t_1)-h^eff(t_1 ) ( t_1)(t)h^l(t ) + & & + ( t_1)(t)h^l(t)h^eff(t_1 ) + & & + _
dttr_l h^eff(t_1)h^all(t)(t)(t)h^all(t ) + & & - _ t^t_1 dt_t^t_1 dttr_l h^all(t)(t)(t)h^all(t)h^eff(t_1 ) + we note that the above equation is a finite - order expansion unlike the traditional dyson series which extends to infinite orders .
the third order terms in @xmath20 are denoted by @xmath158 $ ] .
after some algebraic rearrangements we have from equation ( [ am2 ] ) , [ am3 ] _
t_1(tr_l(t_1)(t_1 ) ) & = & -tr_l-_t^t_1 dttr_lh^eff(t_1)h^eff(t)(t)(t_1 ) + & & -h^eff(t)(t)(t_1)h^eff(t_1)-h^eff ( t_1)(t_1)(t)h^eff(t ) + & & + ( t_1)(t)h^eff(t)h^eff(t_1 ) -_t^t_1
dttr_l + & & + _ t^t_1
dt_t^t_1 dttr_l + & & + _
t^t_1 dt_t^t_1 dttr_l + [ ( h^eff)^3(t ) ] .
averaging both sides of equation ( [ am3 ] ) over the whole ensemble while using the assumptions that @xmath159 is statistically independent of @xmath160 , @xmath161 @xmath162 and @xmath163 we obtain [ am4 ] _
t_1_s(t_1 ) & = & -tr_l-_t^ t_1
dttr_l h^eff(t_1)h^eff(t)(t)(t_1 ) + & & -h^eff(t)(t)(t_1)h^eff(t_1 ) -h^eff(t_1)(t_1)(t)h^eff(t ) + & & + ( t_1)(t)h^eff(t ) h^eff(t_1)+ [ ( h^eff)^3(t ) ] , where @xmath164 = { \rm tr_{\rm l}}[\big\langle\vert\psi(t_1)\rangle\langle\psi(t_1)\vert\big\rangle]$ ] and @xmath165 and @xmath166 denote the density matrices of the system alone and that of the combined system and bath respectively .
following cohen - tannoudji _ et.al . _
we define a coarse - grained version of equation ( [ am4 ] ) for the time - span @xmath24 $ ] , such that @xmath25 , by approximating the instantaneous rate @xmath167 by the coarse - grained rate @xmath168/\delta t$ ] . integrating ( [ am4 ] )
@xmath169 between the limits @xmath24 $ ] and dividing both sides by @xmath32 , while replacing the dummy - variable @xmath170 with @xmath171 , we obtain the coarse - grained master equation @xcite .
[ am5 ] ^cg_t_s(t_1 ) & = & _ s(t + t ) - _ s(t ) + & = & -_t^t + t dt_1tr_l-_t^t + t
dt_1_t^t_1 dt_2tr_l h^eff(t_1)h^eff(t_2)(t_2)(t_1 ) + & & -h^eff(t_2)(t_2)(t_1)h^eff(t_1 ) -h^eff(t_1)(t_1)(t_2)h^eff(t_2 ) + & & + ( t_1)(t_2)h^eff(t_2)h^eff(t_1)+ [ ( h^eff)^3(t ) ] , together with the weak - coupling assumption we also assume that the drive to the system , @xmath26 , is weak such that the terms of third and higher order in @xmath20 become negligible if @xmath32 is much shorter than the time scales associated with @xmath20 ( denoted by @xmath23 ) .
since @xmath172 , we can neglect the time - evolution due to @xmath20 between @xmath14 and @xmath173 in the second - order terms of the coarse - grained version equation ( [ am5 ] ) and also the terms @xmath158 $ ] . assuming that at the beginning of the coarse - graining interval the full density matrix has the factorised form @xmath174 , where @xmath175 denotes the equilibrium density matrix of the bath , @xmath176 being the partition function , we use the usual time ordered propagator @xmath177 due to @xmath19 to obtain [ addm ] ( t_1)(t_2 ) & = & u_l(t_1,t)(t)(t)u_l^(t_2,t ) + & = & _ s(t)_j_j_j-_t^t_1 dtf_j(t ) + _
t^t_2 dtf_j(t ) + & = & _ s(t)_j_j_j(-_j^2t_1 - t_2 ) . in the above expression @xmath178 denotes the eigen - value of @xmath1 corresponding to @xmath179 and we have made use of the fact that @xmath19 and thus the propagators @xmath180 are independent of the initial distribution of the bath states . to obtain expressions in a familiar closed form
we assume that all @xmath8-s have an identical variance , @xmath33 .
thus we obtain [ addm1 ] ( t_1)(t_2)= _ s(t)_l^eq(-^2t_1 - t_2 ) .
substituting ( [ addm1 ] ) in the coarse - grained equation ( [ am5 ] ) , we observe that the second order term takes the form of a double - commutator decaying within the time - scale of @xmath36 .
thus @xmath36 forms the upper bound of the time - scales during which the bath correlations are significant and as such we replace it by @xmath22 . finally , following the prescription of cohen - tannoudji _ et.al .
_ we remove the non - secular contributions by averaging over the coarse - grained time - span , @xmath32 and approximate the coarse - grained derivative of @xmath34 by its ordinary partial derivative to obtain the quantum master equation ( me ) : [ amf ] _
t_s(t ) & = & -_^ + & -&_0^ d_^e^- .
where , the superscript ` sec ' denotes that only secular contributions are retained @xcite .
here we have replaced @xmath32 in the upper limit of the integral in equation ( [ amf ] ) with @xmath181 , since the integrand decays within the short time span @xmath182 , irrespective of the form of the perturbation , @xmath20 @xcite .
_ approach to thermal equilibrium_ to show that me ( [ amf ] ) predicts approach to thermal equilibrium for the system density matrix , @xmath34 , in the absence of an external drive , @xmath26 , we follow the analysis of cohen - tannoudji _ et.al . _ with suitable modifications @xcite . defining the coupling hamiltonian as @xmath183 , ( @xmath184 and @xmath185 are hermitian operators on the relevant and irrelevant subspaces respectively ) we express the me ( [ amf ] ) in the eigen - basis of @xmath0 assuming the bohr frequencies of the system to be non - degenerate @xcite .
the dissipator of the resulting equation involves bath spectral - densities of the form ( ignoring lamb shifts @xcite ) j ( ) & = & _ -^ de^ _ e^- + & = & _ -^ d_j_ke^(+ _
jk)_jk^2e^- + and k(- ) = _
-^ de^- _ e^- , where @xmath186 , @xmath187 , @xmath42 denotes the energy gap ( in angular frequency units ) between any two levels of the system ( bohr frequency ) and @xmath188 is the interaction representation of @xmath185 .
then it suffices to show that @xmath189 @xcite . to this end
we have , [ akms1 ] k(- ) & = & _ -^ d_j_ke^(-+ _ jk)_jk^2e^- + & = & _ j , ke^-_jk_jk ^2 .
+ the lorentzian form thus obtained in the above expression suggests that the maximum contribution to @xmath190 occurs when @xmath191 .
for @xmath192 the width of the lorentzian becomes very small and as such we can replace @xmath193 in the above by @xmath194 and take it out of the summation .
following the usual method of analysis , we then obtain @xmath195 @xcite . on the contrary
when @xmath196 , the lorentzian has an appreciable contribution in the range @xmath197 .
even then the maximum contribution to the spectral densities occurs when @xmath198 . in any case ,
approach to an equilibrium distribution of the system is assured by the me , although it may not always be possible to calculate a closed form expression of that distribution .
to describe the dynamics of the spin @xmath199 system , we assume the zeeman hamiltonian , @xmath200 and the drive , @xmath201 , where @xmath47 is the resonance frequency of the spin-@xmath38 nuclei and @xmath202 , @xmath44 denote the cartesian components of the spin angular momentum operator . since the drive is assumed to be near resonance , we have @xmath203 .
we do not ascribe any specific form to the coupling hamiltonian , @xmath2 but following the usual practice , for an isotropic heat bath , we assume that @xmath204 = 0 $ ] , which in turn ensures that the cross - terms between the drive and the coupling vanish identically in the second - order @xcite .
we consider the effects of the coupling only through relaxation terms proportional to @xmath205 and @xmath206 ( @xmath89 and @xmath90 denote the longitudinal and transverse relaxation times respectively ) @xcite .
a detailed derivation of the relaxation terms will be presented elsewhere .
the perturbative regime , @xmath207 , allows us to neglect the rapidly fluctuating counter - rotating terms of the drive to get @xmath208 $ ] in the interaction picture , where @xmath46 and @xmath209 @xcite .
since the heterodyne detection followed by low - pass filtering in a typical nmr measurement is equivalent to measurements made in a co - rotating frame of frequency @xmath42 , the macroscopic magnetization components are defined by [ ames ] m_(t ) = _ [ _ s(t)e^-t_z_e^t_z ] , where @xmath210 denotes trace over the system degrees of freedom @xcite . with these assumptions , the me ( [ amf ] )
yields the following dynamical equations for the measured macroscopic magnetization components after ignoring any lamb shift terms : where @xmath49 is the equilibrium value of the magnetization .
the resonance offset , @xmath211 is very small in practical cases , so that @xmath212 is not much greater than @xmath132 .
hence the secular approximation retains all the terms in the second - order of the drive .
the dynamical equations ( [ ambl ] ) clearly demonstrate that in the presence of a transverse drive , there exists additional terms quadratic in the drive strength , which in turn lead to faster journey to saturation @xcite .
the equations ( [ ambl ] ) become identical to the bloch equations when @xmath50 is negligible compared to @xmath213 , which is usually the case in most of the solution state nmr experiments . | quantum master equations usually describe the linear response of a driven dissipative system to an external drive . the rabi oscillations and the relaxation from system - bath coupling , as predicted by these equations , forms the basis of all resonant spectroscopy and applications of quantum information processing . here
we report a different formulation of the master equation , describing dynamics beyond the linear response regime , where the drive induces additional damping in the system .
we show that the equilibrium fluctuations of the thermal bath , in the form of gaussian white noise processes , impart a finite memory to the state functions in our iterative expansion of the dynamical equation .
this memory renders the nonlinear response terms time - nonlocal , in a suitably coarse - grained time - scale . in the context of an ensemble of spin qubits ,
we derive a modified form of bloch equations having the leading order non - linear response terms and verify their existence experimentally . since fluctuations are ineradicable in physical systems , our results indicate that manipulation of a qubit ensemble with external drives introduces unwanted damping , resulting in further loss of information .
pacs numbers : : 05.30.d , 05.70.ln , 05.40.-a , 76.60.-k , 03.67.a |
the number of strongly lensed sources found in individual observational programs can be used to constrain the cosmological constant ( e.g. , turner 1990 ) . using statistics of optical lensing programs ,
various studies have now derived upper limuts on the present day value of the cosmological constant ( kochanek 1996a ; chiba & yoshi 1997 ; cooray et al .
as discussed in the literature , however , lensed source search programs at optical wavelengths can be affected by systematics arising from effects due to extinction and reddening ( e.g. , malhotra et al .
1996 ) , and issues related to completeness ( kochanek 1991 ) .
such effects , especially extinction for which there is no clear consensus , can not easily be quantified in cosmological studies with gravitational lensing statistics . at radio wavelengths ,
neither of these effects are likely to affect the statistics ( see , falco et al .
1998 ) , including completeness , when flux - limited systematic surveys are carried out . also , high resolution interferometers such as the vla and merlin allow detection of lensed sources with image separations much smaller than that allowed by ground - based optical telescopes .
thus , statistics from lensed radio source surveys are generally preferred over the optical ones to obtain cosmological parameters
. recently , falco et al . (
1998 ) studied the statistics of lensed radio sources in the jodrell bank - vla astrometric survey ( jvas ; browne et al .
1997 ; patnaik et al .
1992 ) sample of 2500 flat - spectrum sources brighter than 200 mjy at 5 ghz .
they were able to derive a 2@xmath15 lower - limit on the mass density of the universe of @xmath16 ( @xmath17 at 1@xmath15 ) .
recently , the number of strongly lensed radio sources has steadily increased with results from the cosmic lens all sky survey ( class ) , which has observed @xmath0 6500 flat - spectrum sources during the first two sessions of its observations using the vla . the total number of currently confirmed lensed radio sources from these observations amount to six : b1608 + 656 ( myers et al .
1995 ) , b2045 + 265 ( fassnacht et al .
1998 ) , b1933 + 503 ( sykes et al .
1998 ) , b0712 + 472 ( jackson et al .
1998a ) , b1600 + 434 ( jackson et al . 1995 ) , b1127 + 385 ( koopmans et al .
the class is a collaboration between groups at jodrell bank , caltech , dwingeloo and leiden , and is an extension to the previous jvas by observing flat - spectrum sources ( @xmath18 ) down to a flux density limit of 30 mjy at 5 ghz . here
, we calculate the expected number of lensed radio sources in the class sample of @xmath0 6500 sources for different cosmological parameters , and constrain these parameters by comparing the predictions with the observations . in sect .
2 , we discuss our calculation and its inputs , including the redshift distribution for background radio sources and the luminosity distribution of foreground lenses . in sect .
3 , we present our resulting constraints on cosmological parameters , and in particular on @xmath1 , as well as on the redshift distribution of the faint radio sources . in sect .
4 , we discuss the potential effect of systematic errors in the predicted and observed rate of strong lensing . finally , in sect . 5 , we summarize and discuss future prospects for tighter constraints .
we follow the conventions that the hubble constant , @xmath19 , is 100@xmath20 km s@xmath21 mpc@xmath21 , the present normalized mean density of the universe @xmath22 is @xmath23 , and the present normalized cosmological constant @xmath3 is @xmath24 . in a flat universe , @xmath4 .
we have simplified the notation by not including the subscript 0 usually taken to denote the present day values of these quantities .
our present calculational method is quite similar to that of cooray et al .
( 1998a ; hereafter cqm ) , in which we calculated the expected lensing rate in the hubble deep field ( hdf ; williams et al .
1996 ) , as implied by various photometric redshift catalogs for hdf sources . a comparison between the observed number of lensed sources in the hdf and the predicted number , allowed us to calculate the present value of @xmath1 .
we follow a procedure similar to cqm by first establishing the redshift distribution of the observed class sources , and then by using the standard lensing calculations to estimate the expected number of strongly lensed radio sources present in this sample . in fig . 1
, we show the flux density distribution of class sources . except for a handful of sources with flux densities
greater than 200 mjy , most of the observed sources follow a distribution that sharply rises around 30 mjy .
this is only an observational artifact due to the 30 mjy cutoff imposed on the initial source selection . since the source sample was selected based on previous green bank 5 ghz surveys ( e.g. , gregory et al .
1996 ) and that we are now using the measured flux densities from class vla snapshot maps at 8.4 ghz , the presence of sources with flux densities less than 30 mjy at 8.4 ghz suggest that the current class sample includes a small number of non - flat spectrum radio sources ; the class definition with respect to flat spectrum sources is based on an upper frequency point at 5 ghz rather than 8.4 ghz .
sources with flux densities less than 30 mjy are also expected given uncertainties in the green bank survey flux density measurements and the possibility for intrinsic flux density variation between green bank and class survey epochs .
the total number of radio sources in the publicly available catalog of class sourcesnjj / glens / class.html ] during the first two sessions is 6528 . as discussed in falco et al .
( 1998 ) , one of the main problems associated with lensed radio source statistics is the non - availability of redshifts for radio sources .
however , it has also been shown by falco et al .
( 1998 ) , and previously by kochanek ( 1996b ) , that the radio sources exhibit a strong correlation between their redshift and flux density distributions .
thus , even though complete redshift surveys may not exist , it is still possible to construct with existing data a statistically accurate redshift distribution for a given radio flux density .
this assumes that the correlation between redshift and flux density distribution remains to be valid at flux levels lower than observationally studied .
since redshifts can not be determined exactly for individual sources , each of the assigned redshifts at a given flux density will have an uncertainty which can be accounted when calculating cosmological parameters .
also , since the class radio source sample is substantially large , on average , the modeled redshift distribution can be expected to represent the true redshift distribution . since
no new radio source redshift surveys have been published recently , we use the parameterization of the flux@xmath25redshift relation in falco et al .
( 1998 ) to calculate the probable redshifts of class sources . for each redshift
, we also assign an uncertainty based on the error in the parameterization presented by falco et al .
( 1998 ; see , their fig .
since no redshift surveys exist for radio sources with flux densities less than @xmath0 200 mjy , we use the extrapolation of the flux@xmath25redshift distribution to estimate the redshifts of these sources .
however , we assign a slightly higher uncertainty for these redshifts than expected from the simple extrapolation presented in falco et al .
we assume that the model a in falco et al .
( 1998 ) is the mean redshift distribution , while models b and c , with an increase in uncertainty , represents our higher and lower redshift distributions , respectively . with redshift and flux density for each source , we calculate the radio luminosity , @xmath26 , and its uncertainty .
we do not consider any errors in the flux density measurement as these are likely to be small compared to the ones associated with the lensing calculation . in order to calculate the lensing rate for radio sources , we model the lensing galaxies as singular isothermal spheres ( sis ) and use the analytical filled - beam approximation ( e.g. , fukugita et al .
1992 ) . at redshifts
@xmath27 , the analytical filled - beam calculation in fukugita et al . (
1992 ) agree to better than 2% with numerical calculations ( e.g. , fig . 1 in holz et al .
1998 ) . gravitational lensing statistics are affected by the so - called `` magnification bias '' ( kochanek 1991 ) in which the number of lensed sources in the sample can be different from an unlensed sample down to the same flux density level .
thus , any calculation involving lensed source statistics should account for the magnification bias and associated systematic effects .
the effect of this bias , whether to increase or decrease the number of lensed sources , depends on the slope of the number counts . the magnification bias is particularly pronounced in quasar lensing surveys ( maoz & rix 1993 ) , because the faint end of the quasar luminosity function rises steeply , increasing the number of lensed sources . in sect .
2.2 , we discuss this effect for radio sources based on the flat - spectrum radio luminosity function .
following cqm , if the probability for a source at redshift @xmath28 to be strongly lensed is @xmath29 , as calculated based on the filled - beam equation , we can write the number of lensed sources , @xmath30 , as ( see , also , maoz et al . 1992 ) : @xmath31 where @xmath32 is the magnification bias for a radio source at redshift @xmath28 with luminosity @xmath26 , and @xmath33 is the selection function defined based on image separation , @xmath34 , and flux ratio , @xmath35 , between components of the strongly lensed sources . here
, the sum is over each of the radio sources in our sample .
the index @xmath36 represents each object ; hence , @xmath37 and @xmath38 are , respectively , the redshift , and the rest - frame luminosity of the @xmath36th source . for simplicity ,
following kochanek ( 1996b ) , we assume that the selection function is simply a uniform function in the range @xmath39 , such that all lensed sources in this range are recovered . for the sis model , the fraction of lensed sources in this range of image separations is 0.917 ( e.g. , kochanek 1996b ) .
we have assumed that the image separation , @xmath34 , under the sis assumption for foreground galaxies is given by @xmath40 , where @xmath15 is the velocity dispersion of the foreground lensing galaxies , and that the characteristic velocity dispersion for a @xmath41 galaxy , @xmath42 , is 220 km s@xmath21 ( see , eq .
22 in kochanek 1996b ) .
given that the observations are carried out to a much lower flux density level than the flux density of the observed source , and that the candidate lensed sources are selected when flux ratios between strong components are @xmath43 10 ( browne et al .
1997 ) , it is likely that the selection function is independent of the flux ratio between lensed images . for sis models , the fraction of lensed sources with image flux ratios @xmath43 10 is 0.9997 , and thus , for the purpose of this calculation we take @xmath33 to be a constant with numerical value of 0.916 .
the bias , @xmath44 , can be calculated based on the radio luminosity function for background sources at redshift @xmath28 , @xmath45 : @xmath46 where @xmath47 is the probability distribution of amplifications @xmath48 .
the ghz radio luminosity function for flat - spectrum sources , and its evolution , is still not well determined observationally . for the purpose of the present paper ,
we follow the work of dunlop & peacock ( 1990 ) , where luminosity functions are calculated for sources at 2.7 ghz .
since we are dealing with a sample of flat spectrum sources , we can safely assume that the 1 to 10 ghz luminosity function is equivalent to that at 2.7 ghz .
we use the pure luminosity - evolution model ( ple ) for flat - spectrum sources , which is motivated by optical quasar and x - ray agn luminosity functions ( e.g. , boyle et al .
1987 ) : @xmath49^{-1},\ ] ] where @xmath50 and @xmath51 are the faint and bright end slopes of the luminosity function with respect to the critical , or break , luminosity @xmath52 .
for the purpose of this paper , the quantity @xmath26 is the radio luminosity ( in w hz@xmath21 sr@xmath21 ) . since this luminosity function is defined in terms of an element @xmath53 instead of the @xmath54 required in eq . 2 , following king & browne ( 1996 ) , we rewrite the luminosity function in terms of a linear element , with new slopes @xmath55 and
@xmath56 below and above the critical luminosity : @xmath57^{-1}.\ ] ] the critical luminosity , @xmath52 , has been found to evolve with the form : @xmath58 in table 1 , we list each of the coefficients .
cc parameter & value + log(@xmath59/mpc@xmath60 log l@xmath21 ) & -8.15 + @xmath50 & 0.83 + @xmath51 & 1.96 + @xmath55 & 1.83 + @xmath56 & 2.96 + @xmath61 & 25.26 + @xmath62 & 1.18 + @xmath63 & -0.28 + since we are using the sis model , the minimum amplification , @xmath64 , is 2 and the probability distribution for amplifications , @xmath65 , is @xmath66 ( see , e.g. , maoz & rix 1993 ) .
the magnification bias can be written as ( maoz & rix 1993 ) : @xmath67 in fig . 2
, we show the magnification bias as a function of radio luminosity using the luminosity range for most of the sources in the class survey . for sources with luminosities such that @xmath68 , the magnification bias is simply determined by the slope @xmath55 and
is numerically equivalent to @xmath0 3.04 . the probability of strong lensing depends on the number density and typical mass of lensing galaxies . for singular isothermal spheres ,
this factor is conveniently represented by the dimensionless parameter ( turner , ostriker & gott 1984 ) : @xmath69 here @xmath70 is the number density of galaxies , @xmath71 , and @xmath15 is the velocity dispersion .
the parameter @xmath72 is independent of the hubble constant , because the observationally inferred number density is proportional to @xmath73 .
we can estimate @xmath72 at a given redshift from the galaxy luminosity function , which we describe based on the schechter function ( schechter 1976 ) , in which the comoving density of galaxies at redshift @xmath28 and with luminosity between @xmath26 and @xmath74 is : @xmath75^{\alpha(z ) } e^{-l / l^*(z)}\ , dl\;.\ ] ] in order to relate velocity dispersion , @xmath15 , with luminosity , we assume a dependence between absolute magnitude , @xmath76 , and @xmath15 of the form : @xmath77 which is known as the faber - jackson relation for early - type galaxies ( faber & jackson 1976 ) and the tully - fisher relation for spiral galaxies ( tully & fisher 1977 ) .
thus , using eqs . 8 and 9 , eq .
7 can be written as ( fukugita & turner 1991 ) : @xmath78 where @xmath79 is the normal gamma function and @xmath42 is given by the characteristic magnitude @xmath80 : @xmath81 . to estimate @xmath72
, we use a recent determination of the foreground galaxy luminosity function by zucca et al .
( 1997 ) based on observations of 3342 galaxies in the eso slice project ( esp ) .
a comparison between the esp luminosity function and the ones previously used in lensing statistics calculations ( e.g. , kochanek 1996 , falco et al . 1998 ) suggest no large variations .
thus , we expect our @xmath72 parameter to be within the uncertainty of previous @xmath72 parameter estimations . however , we note one change in our estimate for @xmath72 from a recent study of lensed source statistics : we consider lensing galaxies to be composed of both early - type as well as spiral galaxies .
this is contrary to the recent suggestion by chiba & yoshi ( 1998 ) that spiral and dwarf - type galaxies do not contribute to the observed lensing rate .
we argue against this possibility simply based on the observational data , where spiral galaxy lenses have been found ( e.g. , b1600 + 434 in the present class sample , jaunsen & hjorth 1997 ; koopmans et al . 1998 ) .
zucca et al .
( 1997 ) find that the luminosity function for galaxies is represented by a schechter function with parameters : @xmath82 these parameters agree with the recent estimation of the luminosity function of known lenses ( kochanek et al .
1998 ; see their fig . 4 ) . in order to derive the characteristic velocity dispersion @xmath42 for the observed @xmath80
, we use the following faber - jackson and tully - fisher relations : @xmath83 as presented by de vaucouleurs & olson ( 1982 ) for early - type galaxies ( e / s0 ) and fukugita et al .
( 1991 ) for spiral galaxies ( sp ) .
to estimate the overall @xmath72 from individual contributions , we use the galaxy fractions presented by postman & geller ( 1984 ) with ratios e : s0:sp=12@xmath842:19@xmath844:69@xmath844 .
these ratios also agree with the relative numbers of spiral and early type galaxies in the southern sky redshift survey ( marzke et al .
1998 ) . in table 2 , we list these parameters and estimate for @xmath72 using the luminosity function from the esp survey .
cccc type & fraction & @xmath42 & @xmath72 + e & 0.12 @xmath84 0.02 & @xmath85 & 0.009 @xmath84 0.004 + s0 & 0.19 @xmath84 0.04 & @xmath86 & 0.010 @xmath84 0.003 + sp & 0.69 @xmath84 0.04 & @xmath87 & 0.006 @xmath84 0.004 + total & & & 0.026 @xmath84 0.006 + the best estimate for @xmath72 is 0.026 @xmath84 0.006 , with a relative contribution of 31% from spiral galaxies .
such a spiral galaxy contribution agrees with the estimate by helbig ( 1998 ) that 2 of the 6 known lensed class sources are due to foreground spirals .
this observation also justifies our inclusion of spiral galaxies in the @xmath72 parameter estimation .
our estimate for @xmath72 is a factor of 2.5 higher than the value used by chiba & yoshi ( 1998 ) . according to kochanek et al .
( 1998 ) , the two luminosity functions used by chiba & yoshi ( 1998 ) are the two models most discrepant with the luminosity function from known lenses . using a low @xmath72 value can decrease the number of expected lensed sources enhancing the significance of a cosmological constant in producing strongly lensed sources . in order to obtain reliable estimates of the cosmological parameters based on statistics of lensed sources , as well as for comparative purposes ,
a consistent value of @xmath72 is needed among different studies .
we note that our estimate for @xmath72 is consistent with previous estimates by kochanek ( 1996a,1996b ) and falco et al .
since there could be additional errors , we allow for a slightly higher uncertainty ( 30% ) and take @xmath72 to be 0.026 @xmath84 0.008 .
this estimate for @xmath72 is factor of four less than the value we recently estimated for gravitational lensing in the hdf ( cqm ) .
this was both due to the factor of 2 higher number density of sources in the hdf as suggested by its luminosity function ( sawicki et al .
1997 ) and our partly incorrect assumption that all such sources are elliptical galaxies .
the revised constraints on cosmological parameters from hdf lensing rate based on a more appropriate @xmath72 value could be found in cooray et al .
( 1998b ) .
we calculate the expected number of lensed sources by first estimating the redshift , luminosity and lensing rate for individual class sources as a function of cosmology , with luminosity used to account for the magnification bias .
we also vary the redshift according to flux@xmath25redshift correlation and recalculate the expected number of strongly lensed sources . in fig . 3
, we show the expected number of strongly lensed radio sources in the class survey as a function of @xmath2 and @xmath3 when the mean redshift is considered : a universe dominated with @xmath3 has a higher number of multiply - imaged sources than a universe dominated with a large @xmath2 . as shown in fig . 3
, @xmath30 is essentially a function of the combined quantity @xmath1 , which is mainly due to a coincidence ; the exact dependence is a function of background source redshifts . for background source redshifts in the range 1 to 3 , the curves are essentially @xmath1 .
we can use this degeneracy in the lensing rate ( e.g. , carroll , press & turner 1992 ; kochanek 1993 ) to constrain @xmath1 rather than @xmath2 or @xmath3 individually . in table 3 , we list the expected number of strongly lensed radio sources along the @xmath4 line as a function of @xmath1 .
the error bar accounts only for the uncertainty in redshift .
we account for the error in @xmath72 when cosmological parameters are estimated , which is facilitated by the fact that the number of expected lensed sources is directly proportional to @xmath72 .
cc @xmath1 & @xmath30 + -1.0 & @xmath88 + -0.8 & @xmath89 + -0.6 & @xmath90 + -0.4 & @xmath91 + -0.2 & @xmath92 + 0.0 & @xmath93 + 0.2 & @xmath94 + 0.4 & @xmath95 + 0.6 & @xmath96 + 0.8 & @xmath97 + 1.0 & @xmath98 +
as listed in sect .
1 , the class survey has produced six new lensed sources during the first two sessions of its observations from a sample of 6500 sources .
we assume this number to be our canonical case .
however , to account for the possibility that a subsample of lensed sources has been missed , possibly due to unaccountable selection effects and other biases in the lens search strategy , we consider the possibility that there are total of ten strongly lensed sources within this sample and study the variation in cosmological parameters from the canonical case .
we also study the case when there are four lensed sources .
we follow cqm to constrain the quantity @xmath1 by comparing the observed and predicted number of lensed radio sources .
we adopt a bayesian approach , and take a uniform prior for @xmath99 between -1 and + 1 , since we do not yet have a precise determination of this quantity , and because we do not wish to consider cosmologies in which either @xmath2 or @xmath3 lie outside the interval [ 0,1 ] . for values outside the interval of -1 to + 1 in @xmath1 , we assume an a priori likelihood of zero
. we do not constrain @xmath2 or @xmath3 separately ; thus , no prior is required for these quantities . since the prior for @xmath99 is uniform , the posterior probability density is simply proportional to the likelihood . as derived in cqm , the likelihood @xmath100
a function of @xmath99 is the probability of the data , given @xmath101 .
the likelihood for @xmath102 observed sources ( at redshifts @xmath103 ) when @xmath30 is expected is given by : @xmath104\right ) \ ; .\ ] ] here , @xmath102 is the observed number of strongly lensed radio sources , while @xmath30 is the expected number of lensed sources for a given cosmology .
we have also taken into account the uncertainty in @xmath72 by defining @xmath105 and taking @xmath106 to have a mean of unity and standard deviation @xmath107 allowing for a 30% uncertainty in @xmath72 .
the factor @xmath106 is then an overall correction to the expected lensing rate , due to a systematic uncertainty in @xmath72 . in order to constrain @xmath1
, we also need the redshifts @xmath103 of the observed lensed sources . using table 1 of jackson et al .
( 1998b ) , we obtained redshifts for 4 sources .
for the 2 sources with no known redshifts , we assume a source redshift of 1.5 , consistent with other four redshifts as well as the predicted redshift distribution for radio sources in the class sample .
changing these redshifts to reasonably different values do not change our constraints on the cosmological parameters greatly .
we assume that there are 4 , 6 and 10 lensed sources in the class sample .
this is primarily to study the variation in cosmological parameters with the observed number of lensed sources .
cc case & @xmath99 + n=6 , @xmath108 & -0.29 + n=6 , @xmath109 & -0.58 + n=6 , @xmath110 & -0.23 + n=10 , @xmath111 & -0.58 + n=10 , @xmath109 & -0.90 + n=10 , @xmath110 & -0.52 + n=4 , @xmath111 & -0.08 + n=4 , @xmath109 & -0.28 + n=4 , @xmath110 & -0.04 + in table 4 , we summarize our 95% confidence lower limits on @xmath1 for the three cases , with @xmath111 , @xmath109 , and @xmath110 representing the mid , low and high values for the redshift distribution of radio sources . the cumulative probabilities for observing six lensed sources with the three different estimates of the redshift distribution is shown in fig . 4 .
as shown and tabulated , our lowest 95% confidence on @xmath1 is -0.58 .
accordingly , in a flat universe with @xmath4 , @xmath5 ( 95% c.l . ) . if the true redshift distribution for faint radio sources is represented by the high end allowed by the flux@xmath25redshift parameterization , then @xmath112 and @xmath113 ( 95% c.l . ) . if there are ten lensed sources in the present sample of @xmath0 6500 sources in the class survey , assuming that four lensed sources have been missed by current searches , then @xmath114 ( 95% c.l . ) .
these results are consistent with previous lensing statistics with a 95% confidence upper limit on @xmath3 of 0.57 ( kochanek 1996a ) and cosmological parameters based on high redshift type ia supernovae ( riess et al .
however , these limits , especially when @xmath115 , are marginally inconsistent with recent estimates of the cosmological parameters based on the combined cosmic microwave background ( cmb ) power spectrum analysis and the high redshift supernovae ( lineweaver 1998 ; however , see , tegmark 1998 ) , with the current best estimate for @xmath1 of @xmath116 ( 1@xmath15 ) . to be consistent , we need more than the currently known six lensed sources within this class sample or a lower redshift distribution for faint radio sources than allowed by current studies ( @xmath117 , see below ) if there are six or less lensed sources .
the case for a low redshift distribution can be rejected based on the observed distribution of redshifts for known lensed sources . as presented in table 4
, our lower limit on @xmath1 varies widely with the used parameterization of flux@xmath25redshift relation for background radio sources . since the number of sources at the lower flux density end are quite large for the class sample and that the redshift distribution at this low flux density end is not known , it is necessary that the redshift distribution be accurately determined to obtain firm limits on cosmological parameters .
this can be carried out in two ways : observational determination based on spectroscopic observations of optical counterparts of radio sources with flux densities less than 150 mjy or theoretical determinations based on the number counts and other information ( e.g. , jackson & wall 1998 ) .
a comparison of fig . 3 and table 3 , suggest that an upper limit to @xmath1 can also be calculated , in addition to the lower limit . since the number of lensed sources does not vary greatly when
@xmath118 , such an upper limit is weak compared to constraints from other cosmological probes .
also , this upper limit from lensing lies outside the current range of interest in @xmath1 of -1 to + 1 , and will depend on what one assumes for a prior likelihood .
also , flat cosmologies where @xmath1 is greater than + 1 are incompatible with current constraints on cosmological parameters ( e.g. , white 1998 ) .
therefore , we do not attempt to calculate an upper limit to @xmath1 , hence a lower limit to @xmath3 in a flat universe , using current lensing statistics . since the class survey is expected to observe a total of @xmath0 10000 sources , it is possibile that the current constraints on cosmological parameters may be improved . in order to investigate this possibility
, we used the full catalog of @xmath0 10500 sources publicly available from the class collaboration , which includes sources that are planned to be observed , as well as the @xmath0 6500 sources in the current sample .
the present constraints on @xmath1 do not vary greatly when the true number of lensed sources for the whole sample remains to be what is obtained based on the current observed lensing rate .
if for some reason , the additional sample of @xmath0 3500 sources were to contain twice as more lensed sources than currently expected , then cosmological parameter estimates change to be consistent with a high @xmath3 .
however , since this is not likely to be the case , unless the source selection process is heavily affected systematics , an increase in the source sample is not likely to improve the radio lensing constraints on cosmological parameters greatly .
recent studies have suggested the existence of a scalar field as viable alternatives to the cosmological constant ( see , e.g. , caldwell , dave & steinhardt 1998 ) with an equation of state of the form @xmath6 , where @xmath7 and @xmath8 are the pressure and density of the scalar - field , respectively .
then , as the universe expands @xmath119 , where @xmath120 is the scale factor with @xmath121 . the vacuum energy or the cosmological constant corresponds to @xmath122 , while texture or tangled strings correspond to @xmath123 .
we modified our lensing probability calculation to account for the scalar field , and recalculated the likelihoods using the same technique as above .
we assume a flat universe , and take @xmath11 , where @xmath10 is the present day normalized energy density of th scalar - field component .
we assume a prior in which @xmath10 is uniform between 0 and + 1 .
our new constraints are summarized in fig . 5 on the @xmath9 plane .
the 95% and 68% upper limits ( _ thick solid lines _ ) from gravitational lensing are labeled 95% gl and 68% gl respectively . the 95% confidence . upper limit on @xmath10 can be written as @xmath124 .
in addition to lensing constraints , we also show the resulting constraints on @xmath10 and @xmath125 based on type ia supernovae at high redshifts ( 95% sn ) as recently studied by garnavich et al .
( 1998 ; _ dot - dashed lines _ ) and constant age lines of the universe as a function of @xmath126 ( _ dotted lines _ ) . for a hubble constant , @xmath19 , of 65 @xmath84 10 km s@xmath21 mpc@xmath21 ( riess et al . 1998 ) and a minimum age of the universe , @xmath127 , from globular clusters of @xmath128 gyr ( chaboyer et al .
1996 ) , a conservative lower limit on @xmath126 is @xmath129 ( _ dashed line _ ) . for an age of the universe of @xmath0 15 @xmath84 2 gyr , @xmath130 0.90 to 1.15 .
these three cosmological probes , type ia supernovae , age and lensed sources , qualitatively agree with each other when @xmath131 and @xmath132 .
these combined results rule out tangled strings with an equation of state of the form @xmath123 as the scalar - field component .
as studied by turner & white ( 1997 ) , the combined large scale structure , age and supernovae data are consistent with @xmath133 when @xmath134 to 0.4 .
the constraints on @xmath2 primarily comes from the clustered mass - density in galaxy clusters based on observed baryonic mass combined with a nucleosynthesis determined baryonic mass density of the universe , @xmath135 ( e.g. , burles & tytler 1997 ) , resulting in @xmath136 ( evrard 1997 ) . in fig . 5
, we have outlined a conservative lower limit on the known matter content considering these and similar measurements .
even with known matter considered , there is still a large region left on the @xmath9 plane to form a flat universe allowing various possibilities for the physical nature of the scalar - field component . in future ,
constraints on the equation of state of the unknown scalar - field component is likely to be improved based on cmb anisotropy and large scale structure data combined with standard cosmological probes such as high redshift supernovae and gravitational lensing statistics ( see , hu et al .
1998 ) .
since redshifts for radio sources with flux densities less than 150 mjy are not accurately known , we can use the observed lensing rate with an assumed cosmological model to constrain the redshift distribution of background sources .
this approach is very similar to the one used by holz et al .
( 1998 ) to constrain the redshift distribution of gamma - ray burst sources detected by batse based on lensing statistics .
for the purpose of this calculation , we vary the redshifts of the background sources with flux densities less than 150 mjy , while keeping the rest fixed at the appropriate redshifts as determined earlier .
since we are only interested in an upper limit to the average redshift of faint radio sources , we calculate lensing probabilities as a function of @xmath13 , the _ effective _ average redshift , under the assumption that all sources are at this redshift .
this methodology parameterizes our ignorance of the faint radio source redshift distribution ( see , holz et al .
1998 for a discussion ) .
however , this assumption is incorrect given that at a given flux density level , radio sources have a rather broad distribution in redshift .
nevertheless , our approach is to suggest a possible upper limit to the mean redshift of this distribution , assuming that the distribution is not shaped such that more sources are at higher redshifts than suggested by the mean .
since the lensing rate is highly nonlinear in redshift , a slowly decreasing distribution with increasing redshift can increase the lensing rate faster than the assumed average value . in general
, our upper limit underestimates the true average redshift . in table 5 , we tabulate the 68% confidence upper limit on the redshift of the subsample of sources with flux densities less than 150 mjy and assuming that the total number of lensed sources in the current class sample is either 6 or 10 .
this upper limit on the redshift is again a function of @xmath1 .
if there are 6 lensed sources in the present class sample , then @xmath137 ( 68% c.l . ) . since the observed redshift distribution of known lensed sources in the class sample are between 1.3 and 1.6 ( see , table 1 in jackson et al .
1998b ) , cosmological models where @xmath3 is greater than 0.8 in a flat universe are disfavored .
ccc @xmath1 & @xmath115 & @xmath138 + -1.0 & 0.57 & 0.84 + -0.4 & 1.05 & 1.39 + 0.0 & 1.44 & 1.78 + 0.4 & 1.83 & 2.24 + 1.0 & 2.37 & 2.88 +
as presented in table 4 , a major uncertainty comes from the unknown redshift distribution of faint radio sources with flux densities less than 150 mjy .
thus , it is unlikely that the current observational data on radio sources would be useful to constrain the cosmological parameters based on class results , unless the redshift distribution for its sources is properly established .
since the redshift and flux density distributions for radio sources are strongly correlated , it is not necessary that redshifts for all @xmath0 6500 sources be known .
however , redshifts for a representative subsample of class sources are much needed .
the determination of redshifts would involve the process of optical identifications and spectroscopic observations ; a task that would take a considerable amount of telescope time and man power to complete .
however , as a result of this study , where we have shown the strong variation in cosmological parameters with a change in redshift distribution , it is necessary that the true redshift distribution be used to constrain the cosmological parameters .
we strongly recommend that a subsample of sources be defined from the class catalog and be followed up optically to obtain redshifts .
the determination of redshifts would also aid in the estimation of a reliable luminosity function for faint radio sources , which can be used to improve the calculation associated with the magnification bias .
in addition to redshift distribution , the selection function for lensed radio source discovery process is badly needed .
for the present calculation , we have assumed that all lensed sources with image separations between 0.3 and 6 arcsecs are recovered .
however , the selection function is likely to be complicated ; it is much easier to recognize a four - image lensed source than a doubly - imaged lensed source .
an initial study of such biases as applied to the jvas survey could be found in king & browne ( 1996 ) .
however , as discussed earlier , since radio lens statistics are much cleaner than the optical ones , we do not expect biases and selection effects to make a big difference in present constraints on the cosmological parameters , unless the selection function is heavily affected by such effects .
nevertheless , king & browne ( 1996 ) found that the jvas survey contains at least a factor of 2 less doubly imaged sources than expected from sis and external shear models .
such a result , if also confirmed for class survey , suggests the possibility that a large number of doubly - imaged sources has been missed by the search process or that a bias not yet understood affects the observations .
apart from background sources , an understanding of the foreground lenses is clearly needed .
use of different @xmath72 values by neglecting certain galaxy populations can lead to rather diverse results , as was recently demonstrated by chiba & yoshi ( 1998 ) .
current deep and wide - area redshift surveys are likely to increase our knowledge on the foreground lenses and their composition over the next few years . a consistent description of foreground lenses are also needed for the comparative purposes between optical and radio lensed source sample studies . currently , however , this task is complicated by individual studies suggesting different values for the @xmath72 parameter , resulting from different luminosity functions used to describe foreground lenses .
with increasing knowledge on the foreground lenses , an ideal approach would be to use the luminosity function determined from known lensing galaxies ( e.g. , kochanek et al . 1998 ) , including the variation of their number density with redshift .
using flux density distribution for a sample of @xmath0 6500 radio sources observed during the first two sessions of the cosmic lens all sky survey , and the observed number of strongly lensed sources within this sample , we have constrained the cosmological parameters .
we have considered the possibility that there are total of six ( as observed ) , four or ten lensed sources within this sample to investigate the changes in our constraints on the cosmological parameters .
we find that the expected number of lensed sources is primarily a function of @xmath1 .
a comparison of the predicted number of lensed sources with the observed number allows us to constrain its current value .
based on the current determinations of the redshift distribution for radio sources , our 95% confidence lower limit on @xmath1 is between -0.58 to -0.23 depending on the assumed model for the redshift distribution of faint radio sources .
for a flat universe with @xmath4 , then , @xmath113 to 0.79 ( 95% c.l . ) .
if there are ten strongly lensed sources , the 95% confidence lower limit on @xmath3 is between -0.90 and -0.52 .
these limits are in agreement with previous limits based on gravitational lensing ( kochanek 1996a , 1996b ; falco et al .
1998 ) , and are not in conflict with estimates based on high redshift supernovae ( viz . , @xmath139 [ riess et al . 1998 ] ) .
as has been performed recently in various papers , combining @xmath1 results from high redshift supernovae measurements with @xmath140 results from cmb power spectrum analysis constrains @xmath2 and @xmath3 separately , with much higher accuracy than the individual experiments alone .
we note that gravitational lensing constraints on @xmath1 should also be considered in such an analysis , to obtain a consistent picture on the present cosmological parameters .
in addition to the cosmological constant , we have also studied the possibility of a scalar field responsible for the additional energy density of the universe .
we have shown our constraints on the @xmath9 plane .
if there are six strongly lensed sources in the present class sample , gravitational lensing statistics allow us to rule out the region with @xmath12 ( 95% c.l . ) .
we study the region allowed using combined constraints from gravitational lensing statistics , high redshift type ia supernovae distances and limits on the age of the universe from globular clusters .
we have also considered the current estimates on the known matter content of the universe .
since we have assumed a redshift distribution for radio sources to constrain cosmological models , alternatively , we can constrain the redshift distribution of faint radio sources based on the observed gravitational lensing rate and an assumed cosmological model .
if there are six strongly lensed sources , the 68% confidence upper limit on the redshift distribution of flat - spectrum radio sources with flux densities less than 150 mjy at 8.4 ghz is @xmath141 . in order to obtain a much tighter estimate on @xmath1 , it is essential that the redshift distribution for radio sources at the faint flux density levels be observationally determined .
also , estimates of the flat - spectrum radio luminosity function and the selection function in lensed source discovery process are needed to complete the story .
we strongly recommend that statistically complete optical spectroscopic programs be carried out to obtain redshifts of faint background radio sources . until such redshifts are obtained , it is unlikely that a major improvement could be made with respect to radio lensing constraints on cosmological parameters .
i thank members of the class collaboration for making their source catalog publicly available , phillip helbig for useful correspondences , and jean quashnock , cole miller , andr fletcher , and john carlstrom for useful discussions .
i would also like to thank the anonymous referee for constructive comments on the paper .
this work was supported in part by the mccormick fellowship at the university of chicago , and a grant - in - aid of research from the national academy of sciences , awarded through sigma xi , the scientific research society . | we calculate the expected number of strongly lensed radio sources in a sample of @xmath0 6500 sources observed with the very large array as part of the cosmic lens all sky survey ( class ) during the first two sessions of its observations . a comparison between the predicted and the observed number of lensed radio sources allows a determination of the current value of @xmath1 , where @xmath2 is the cosmological mass density of the universe and @xmath3 is the normalized cosmological constant .
if there are six strongly lensed sources in this sample , our 95% confidence lower limit on @xmath1 is -0.58 . for a flat universe with @xmath4 , then , @xmath5 ( 95% c.l . ) .
if there are ten strongly lensed sources , the 95% confidence lower limit on @xmath1 is -0.90 .
these lower limits are consistent with estimates based on high redshift supernovae and with previous limits based on gravitational lensing . instead of considering a simple cosmological constant
, we also consider the possibility of a quintessence scalar field responsible for the additional energy density of the universe , with an equation of state of the form @xmath6 , where @xmath7 and @xmath8 are the pressure and energy density of the field .
we present our constraints on the @xmath9 plane , where @xmath10 is the present day normalized energy density of the scalar - field component , assuming a flat universe such that @xmath11 .
if there are 6 strongly lensed sources in the present class sample , gravitational lensing statistics allow us to rule out the region with @xmath12 ( 95% c.l . ) .
we discuss the region allowed by combined gravitational lensing statistics , high redshift type ia supernovae distances , and globular cluster ages . instead of a cosmological model , we can constrain the redshift distribution of faint radio sources based on the observed gravitational lensing rate and an assumed cosmological model . if there are six strongly lensed sources , the 68% confidence upper limit on the average redshift @xmath13 of radio sources with flux densities less than 150 mjy at 8.4 ghz is @xmath14 . in order to obtain a much tighter estimate on the cosmological parameters , it is essential that the redshift distribution for radio sources at the faint flux density levels be observationally determined .
we strongly recommend that statistically complete optical spectroscopic programs be carried out to obtain redshifts for a representative subsample of faint background radio sources . until such redshifts are obtained , it is unlikely that a major improvement could be made with respect to lensed radio source constraints on cosmological parameters . |
@xmath0/@xmath1 + jets is a valuable sample for analysis at the tevatron .
these processes play an important role in the tevatron and lhc physics programs ; @xmath0/@xmath1 + inclusive jets will be a valuable standard model calibration sample at the lhc and @xmath0/@xmath1 + heavy flavor are significant backgrounds to top , higgs and other new physics searches at both the tevatron and lhc .
state - of - the - art leading order ( lo ) and next - to - leading order ( nlo ) calculations on these processes are the focus of several active theory collaborations .
the predictions from these calculations would benefit from experimental verification .
below are described important tevatron results on @xmath0/@xmath1 + inclusive jets , @xmath0 + single @xmath3 and @xmath0/@xmath1 + @xmath2-jets and how these results compare to available theoretical predictions .
events with a @xmath0 or @xmath1 boson and one or more light flavor jets are relatively common at the tevatron .
the high statistics of this sample allows one to probe the validity of predictions from perturbative quantum chromodynamics ( pqcd ) .
nlo calculations for qcd processes are being pursued by several theory collaborations , and the validation of these results are necessary if the nlo predictions are to be widely used . in @xmath0
+ @xmath4 1 jet events in 320 pb@xmath5 of cdf run ii data.,scaledwidth=70.0% ] in @xmath1 + @xmath4 1 jet and @xmath1 + @xmath4 2 jets in 1.7 fb@xmath5 of cdf run ii data.,scaledwidth=65.0% ] the cdf experiment has studied the production of jets in events with @xmath0 and @xmath1 bosons @xcite .
@xmath6 events are selected by identifying a high @xmath7 , central electron along with significant missing transverse energy , @xmath8 ; @xmath9 events are selected by requiring one such electron with another that is either central or in the forward region of the calorimeter , with the invariant mass of the electron pair required to be near the @xmath1 mass peak .
events are then assigned to bins of minimum jet multiplicity .
major sources of background in the @xmath0+jets analysis include events with fake @xmath0 s and electroweak sources ( @xmath10 , single top , dibosons ) ; backgrounds in the @xmath1+jets analysis are dominated by multijet production and @xmath0+jets events in which the @xmath1 signal is faked .
acceptance for these events is studied using simulated signal samples ; the differential cross section for the jets in these events is then examined and compared to some available theory predictions as depicted in figures [ fig : wjets ] and [ fig : zjets ] . from figure
[ fig : wjets ] one can see that the nlo prediction from mcfm @xcite is accurately reproducing the jet @xmath7spectrum in @xmath0 + 1 or 2 jets . for higher multiplicity events ,
lo calculations are necessary .
the current preferred method for generating such events at lo relies on generating multiple samples using a matrix element calculation at fixed orders in @xmath11 and then employing a parton shower program to add in additional soft , colinear jets . matching algorithms
have been designed to identify events that could be double counted in this recipe . from figure
[ fig : wjets ] one can see that the lo prediction consisting of the matrix element calculation from madgraph @xcite , parton shower from pythia @xcite and matching scheme from ckkw @xcite is superior to that of alpgen + herwig shower + mlm - matching @xcite .
it remains to be understood which component of the prediction is causing the difference in these lo predictions . in figure
[ fig : zjets ] one can see that the nlo prediction from mcfm accurately reproduces the jet @xmath12 spectrum in @xmath1+jets events , providing additional confirmation of the validity of the nlo predictions .
@xmath13+single-@xmath3 production is an important process at the tevatron .
@xmath13+single-@xmath3 events are produced via gluon- strange quark scattering , and thus this process offers insight on the strange content inside the proton .
the process also allows an opportunity to measure @xmath14 in a @xmath15 regime not yet probed .
also , @xmath13+@xmath3 contributes to the background to top production and prominent higgs search channels at the tevatron .
cdf @xcite and d @xcite have measured the @xmath13+@xmath3 process in run ii using a similar strategy .
leptonic @xmath13 decays ( @xmath16 with @xmath17 ) are selected via a high @xmath12 isolated central lepton and large @xmath8 . among the required jets in the selected events ,
evidence is sought for semileptonic hadron decay through the identification of a soft muon inside the jet cone .
it is a feature of @xmath13+@xmath3 production that the electric charge of the @xmath13 and @xmath3 are opposite .
the sign of the @xmath3 quark is determined from the charge of the muon used to identify semileptonic hadron decay .
an excess of opposite - sign primary lepton and soft muon events is indicative of @xmath13+@xmath3 production .
opposite sign backgrounds include drell - yan production of @xmath18 , @xmath19 production and fake @xmath13 s .
cdf measured in 1.7 fb@xmath5 of data the production cross section for @xmath13+@xmath3 times the leptonic branching ratio of the @xmath13 , @xmath20 9.8 @xmath21 2.8(stat ) @xmath22 ( syst ) @xmath21 0.6(lum ) pb for events with @xmath23 20 gev/@xmath3 and @xmath24 @xmath25 1.5 .
this can be compared to the nlo prediction from mcfm of 11.0 @xmath26 .
d measured in 1 fb@xmath5 of data the ratio @xmath27 ; measuring the ratio has the virtue that numerous sources of systematic error cancel out .
the result @xmath28 0.071 @xmath21 0.017 is reasonably consistent with a lo prediction from alpgen of 0.040 @xmath21 0.003 .
+ @xmath2-jets analysis in 1.9 fb@xmath5 of data .
[ fig : wbjets],scaledwidth=80.0% ] @xmath0/@xmath1+@xmath2 jet signatures are important backgrounds to top and higgs channels at the tevatron .
separate analyses were undertaken to measure the @xmath2-jet cross section in @xmath0 and @xmath1events with increased precision in the hopes of improving the understanding of these final states .
the event selection for the @xmath0+@xmath2 jets analysis is similar to that employed in the @xmath13+@xmath3 analysis discussed above . here
however @xmath2 jets are selected via the identification of a secondary decay vertex well - separated from the primary @xmath29 interaction point . among the jets possessing vertex tags ,
the @xmath2 content is extracted via a maximum likelihood fit of the vertex mass , which is the invariant mass of the charged particle tracks comprising the secondary vertex .
this variable is discriminant among the different species of jets ; from figure [ fig : wbjets ] one can see that among the tagged jets @xmath30 71% are found to be from @xmath2 .
backgrounds to this @xmath0+@xmath2-jets signal include top production , diboson production and fake @xmath0 s .
signal acceptance was studied with simulated @xmath0+@xmath2-jet events using the alpgen event generator .
signal events are considered from a restricted region of phase space ( @xmath31 with @xmath12 @xmath32 20 gev/@xmath3 , @xmath24 @xmath25 1.1 , a neutrino with @xmath12 @xmath32 25 gev/@xmath3 and exactly 1 or 2 @xmath7 @xmath32 20 gev , @xmath24@xmath25 2.0 jets ) to avoid strong dependence on the signal model in regions where we are not experimentally sensitive . the @xmath2-jet cross section in @xmath0 events in 1.9 fb@xmath5 of cdf run ii data
was measured to be @xmath33 , where the systematic error is dominated by the uncertainty in the vertex mass shape one assumes for @xmath2 jets .
this jet cross section result can be compared to the prediction from alpgen of 0.78pb , a factor of 3 - 4 lower than what is observed in the data .
work is ongoing to understand the difference .
the @xmath1+@xmath2-jet analysis used a similar technique to extract the @xmath2 content of its tagged jet sample .
this analysis has succeeded in examining differential cross sections for the @xmath2 jets in @xmath1events .
one can see that the differential @xmath2-jet cross sections versus jet @xmath12 ( figure [ fig : zbjetpt ] ) and @xmath24 ( figure [ fig : zbjeteta ] ) are not reproduced in all bins by any of the predictions that were constructed .
pythia appears to do a reasonable job at low jet @xmath12 but less so as the jet @xmath12 increases .
the alpgen and mcfm predictions are consistent with each other but not with the data except for a few bins .
it remains to be understood why the predictions are so different .
+ @xmath2 jet differential cross sections as a function of jet @xmath12 from cdf s 2 fb@xmath5 result.,scaledwidth=65.0% ] + @xmath2 jet differential cross sections as a function of jet @xmath24 from cdf s 2 fb@xmath5 result.,scaledwidth=65.0% ]
the @xmath0/@xmath1 + jets samples at the tevatron offer a valuable high statistics testbed for state - of - the - art pqcd calculations .
it appears that for inclusive jet production the nlo predictions are accurately describing the data for @xmath0/@xmath1 + up to 2 jets .
predictions for higher parton multiplicity events at nlo would be beneficial . as for @xmath0/@xmath1 + heavy flavor , nlo predictions for the integrated cross section for @xmath0 + single-@xmath3 appear to be accurate . a consensus on @xmath0/@xmath1
+ @xmath2-jets has yet to be reached ; both lo and nlo predictions do not consistently reproduce the integrated or differential rates of these events in the data .
we thank the fermilab staff and the technical staffs of the participating institutions for their vital contributions .
this work was supported by the u.s .
department of energy and national science foundation ; the italian istituto nazionale di fisica nucleare ; the ministry of education , culture , sports , science and technology of japan ; the natural sciences and engineering research council of canada ; the national science council of the republic of china ; the swiss national science foundation ; the a.p .
sloan foundation ; the bundesministerium fuer bildung und forschung , germany ; the korean science and engineering foundation and the korean research foundation ; the particle physics and astronomy research council and the royal society , uk ; the russian foundation for basic research ; the comision interministerial de ciencia y tecnologia , spain ; the european community s human potential programme under contract hprn - ct-20002 , probe for new physics ; cea and cnrs / in2p3 ( france ) ; fasi , rosatom and rfbr ( russia ) ; cnpq , faperj , fapesp and fundunesp ( brazil ) ; dae and dst ( india ) ; colciencias ( colombia ) ; conacyt ( mexico ) ; krf and kosef ( korea ) ; conicet and ubacyt ( argentina ) ; fom ( the netherlands ) ; stfc ( united kingdom ) ; msmt and gacr ( czech republic ) ; crc program , cfi , nserc and westgrid project ( canada ) ; bmbf and dfg ( germany ) ; sfi ( ireland ) ; the swedish research council ( sweden ) ; cas and cnsf ( china ) ; and the alexander von humboldt foundation . | collider signatures containing bosons and jets are particularly interesting .
recent theoretical effort has been devoted to determining predictions of @xmath0/@xmath1 + multiple parton production ; the high statistics sample of @xmath0/@xmath1 + jets events collected at the tevatron is a valuable testbed for probing the validity of these calculations .
the final state containing a @xmath1 or @xmath0 boson and one or more @xmath2-jets is a promising higgs search channel at the tevatron and could be a window to new physics at the lhc .
these searches benefit from a deep understanding of the production of @xmath0/@xmath1 + heavy flavor jets which constitutes a significant background to the more exotic sources of this signature .
herein the latest tevatron results on these production mechanisms are reviewed with an emphasis on comparison of data results to the latest theoretical models . |
iron - based superconductors ( fesc ) have attracted intense attention since the discovery of superconductivity in lafeaso@xmath5f@xmath6 with t@xmath4 of 26 k. @xcite a central topic of the studies is to understand the mechanism behind superconductivity .
since the electron - phonon coupling strength was proved to be insufficient to explain the relative high t@xmath4 both theoretically and experimentally , @xcite spin fluctuations are expected as an alternative candidate for the pairing glue . therefore understanding the magnetic behavior of the parent compounds
is urgently needed .
unexpectedly , the parent compounds of fesc exhibit a wide variety of magnetic properties despite their similar crystal structures . the undoped compounds are antiferromagnetic metals with the moments ordered in stripes , i.e. ferromagnetically between the nearest fe neighbors in one direction ( x ) and antiferromagnetically in the other ( y ) .
the moment of the 1111 compounds rfeaso ( r = rare earth ) is @xmath1 0.3 - 0.8 @xmath2 .
@xcite in the 122 compounds afe@xmath7as@xmath7 ( a = ca , sr ,
ba , eu ) it is @xmath1 0.8 - 1.0 @xmath2 , @xcite whereas in the 111 family the moment is only @xmath1 0.09 @xmath2 in nafeas @xcite and could not be observed in lifeas .
@xcite in the 11 family , no local moment is found in fese , @xcite while fete has a unique magnetic structure , a so called double - stripe order , with a fairly big moment @xmath1 2.0 @xmath2 .
@xcite this spread in magnetic moments is difficult to understand since the parent compounds have very similar lattice structures ; they all have similar fex ( x = as , se , te ) layers , and density functional theory ( dft ) calculations give quite similar paramagnetic band structures for all the materials .
@xcite it is conceivable that theoretical studies may clarify this puzzle .
dft with the local spin density approximation ( lsda ) or the generalized gradient approximation ( gga ) successfully reproduced the single - stripe antiferromagnetic ( afm ) order in 1111 and 122 materials .
@xcite however the calculated magnetic moment is about 1.5 - 2 @xmath2 , which is much larger than the measured magnetic moment .
many theories have been proposed to explain the outstanding discrepancy between dft and experiment .
yildirim applied a fixed - spin - moment calculation to lafeaso and found the ground state with a moment of @xmath8 @xmath2 .
@xcite nakamura @xmath9 @xmath10 . applied the lsda+u approach to lafeaso@xmath5f@xmath6 and reproduced the experimental moment with a @xmath11 u. @xcite mazin and johannes argued that the small moment is due to the presence of antiphase boundaries .
@xcite bascones @xmath9 @xmath10 . presented a five - band model and found a low moment state characterized by anti - parallel orbital magnetic moments .
@xcite however , it is still not clear which picture is more realistic .
in this work , we present an lsda+@xmath0 study to explain the puzzle of magnetism in fesc .
we show that the measured small magnetic moment can be reproduced with @xmath12 @xmath0 , and its origin is a coulomb - enhanced orbital hybridization between antiferromagnetic neighbors .
since the iron compounds are weak to moderately correlated systems , @xcite the around - mean - field ( amf ) functional @xcite is employed in this work .
all the calculations are performed with the full - potential linear augmented plane wave ( flapw ) within the local - density approximation , as implemented in package wien2k .
@xcite we concentrate on the 1111 material lafeaso and use the experimental lattice structure @xcite with the space group p4/nmm , a=4.035 , c=8.7409 , z@xmath13=0.1415 and z@xmath14=0.6415 . to study the single - stripe afm order , we use the @xmath15 supercell containing 4 iron atoms .
144 k points are used within irreducible brillouin zone .
the muffin - tin sphere radii of 2.24 , 2.28 , 2.02 and 1.98 a@xmath16 are used for la , fe , as and o , respectively .
since lsda+@xmath0 may give multiple minima , @xcite we change the starting magnetic moment on fe to search for all the solutions .
[ mom ] shows our calculated local magnetic moment on fe and the total energy as a function of the coulomb interaction @xmath0 .
we used two values for the hund s coupling , @xmath17=0 and @xmath17=0.82 ev , the latter having been obtained from an x - ray experiment .
@xcite surprisingly , two kinds of low moment ( lm ) in addition to one high moment ( hm ) solutions are found in our calculations .
the hm state was expected since the lsda+@xmath0 is usually believed to have a tendency to enhance the magnetic moment .
the two lm states have the magnetic moments [email protected] @xmath2 ( lm1 ) and [email protected] @xmath2 ( lm2 ) , which are in good agreement with the measured moments in the 122 compounds and in most of the 1111 compounds , respectively .
we did the same calculation for the 122 materials bafe@xmath7as@xmath7 , and also found two lm states , very similar to lafeaso . as shown in fig .
1a and 1b , the lm1 state only exists within a certain @xmath0 range , and it can jump to the hm or lm2 states with a discontinuous change of the magnetic moment when @xmath0 decreases or increases .
the lm2 state requires larger @xmath0 and shows slight variation in moment when @xmath0 is increasing .
fig . 1c and
1d present the total energy of the lm1 and lm2 states relative to the hm state .
the figure shows that a small @xmath0 favors the hm state and a large @xmath0 favors the lm2 state .
consequently , the experimental ground state of lafeaso @xcite can be reproduced in our calculation if @xmath0 is large enough .
the effect of @xmath17 may also be seen in fig .
1 . in general , the results for @xmath17=0 and @xmath17=0.82 ev are similar .
we therefore only consider the case of @xmath17=0 in the following parts of this work .
[ 0.65].,title="fig : " ] our theoretical band structures of lafeaso for the different states are presented in fig .
the lm1 and lm2 states show very different band structures .
the lm1 indicates a rather good metal while the lm2 state is nearly semiconducting .
therefore , the magnetic strength and electrical conductivity are correlated properties in the iron compounds .
the coexistence of the lm1 and lm2 states in our calculations may clarify some seemingly inconsistent experiments .
mcguire @xmath9 @xmath10 .
@xcite reported a systematic study on the 1111 materials .
they found that cefeaso has a magnetic moment about 0.3 @xmath2 and its low temperatures electrical resistivity is @xmath18=2 m@xmath19 cm . while neutron scatting experiment by zhao @xmath9 @xmath10 .
@xcite indicated a magnetic moment of 0.8 @xmath2 for cefeaso , and jesche @xmath9 @xmath10 .
@xcite reported an electrical resistivity of @xmath18=140 @xmath20 cm for cefeaso .
it should be noticed that mcguire @xmath9 @xmath10 .
used polycrystalline samples and jesche @xmath9 @xmath10 . used large single crystals . moreover , a recent neutron scatting experiment @xcite found single crystal lafeaso has a moment of 0.8 @xmath2 , which is also very different from the earlier study .
@xcite from our theory , these discrepancies can be naturally explained by assuming that the samples were in different states , which may be due to the quality of samples .
[ 0.65 ] band structures are for @xmath17=0 and @xmath0=3.26 ev.,title="fig : " ] in the lsda+@xmath0 calculations , the on - site elements of the density matrix may contain some important information about the detail of the states . in table
[ occ ] , we present the occupancy of the fe @xmath21 orbitals and the contribution of each orbital to the spin magnetic moment . the main elements of the spin - dependent local density matrices can be extracted from this table .
the main difference between the three lsda+u states lies in the moment of the @xmath22 orbital . in the hm state ,
@xmath23 is enhanced from the lsda value of 0.42 @xmath2 to 0.52 @xmath2 .
but in the lm1 and lm2 states , @xmath23 is reduced to 0 and -0.43 @xmath2 !
this unusual reduction implies that the @xmath22 orbital is the key to understand the lm states .
the hm , lm1 and lm2 states may be characterized as @xmath23@xmath240 , @xmath23@xmath10 and @xmath23@xmath250 , respectively .
we may therefore name the lm states as orbital - spin ordered states . in the lm2 state , @xmath23 and
@xmath26 have opposite signs from @xmath27 and @xmath28 .
it is worth noting that this result is in very good agreement with the tight - binding study @xcite by bascones @xmath9 @xmath10 .
8.2cmc@c@c@c@c@c@c & @xmath29 & @xmath30 & xy & xz & yz & total + + occupancy & 1.29 & 1.19 & 1.21 & 1.15 & 1.34 & 6.19 + moment & 0.35 & 0.30 & 0.42 & 0.44 & 0.25 & 1.76 + + occupancy & 1.25 & 1.16 & 1.19 & 1.09 & 1.47 & 6.16 + moment & 0.43 & 0.42 & 0.52 & 0.56 & 0.13 & 2.05 + + occupancy & 1.37 & 0.98 & 1.45 & 1.06 & 1.33 & 6.19 + moment & 0.06 & 0.28 & 0.00 & 0.57 & 0.08 & 0.99 + + occupancy & 1.36 & 1.05 & 1.24 & 1.08 & 1.44 & 6.18 + moment & 0.03 & 0.31 & -0.43 & 0.55 & -0.05 & 0.41 + [ occ ] compared with the lsda solution , lsda+@xmath0 rearranges the orbital occupancy and orbital spin - moment .
the lsda+@xmath0 method can be mapped onto a single - band picture if there is no strong orbital hybridization . in this case , the coulomb interaction has two possible effects on the magnetic moment .
the first is to enhance the exchange splitting between the majority spin and minority spin , which causes larger moment for each orbital .
the second is to induce a charge transfer from the less occupied orbitals to more occupied orbitals , which may decrease the moment of the most occupied orbital .
as may be seen , the hm solution agrees well with this picture , but the strong rearrangement of @xmath23 and @xmath31 in the lm solutions can not be understood .
the deficiency of the single - band picture suggests that interorbital coupling plays an important role in the lm states .
this has been realized by bascones @xmath9 @xmath10 .
@xcite they stressed that it is important to include all five @xmath32 orbitals in model studies .
[ fatband ] shows the band character for the @xmath33 and @xmath34 orbitals , where @xmath35 stands for the majority spin and @xmath36 for the minority spin .
one may see that the hybridization of these two orbitals is strongly enhanced in the lm states relative to the lsda solution .
this strong hybridization between different spin channels ( or more precisely , between antiferromagnetic neighbors ) rationalizes the reduction of @xmath23 and @xmath31 in the lm states .
similar enhanced hybridization is also found between some other antiferromagnetic orbitals , such as @xmath37-@xmath34 and @xmath38-@xmath34 .
here we focus on this @xmath33-@xmath34 hybridization .
the strength of orbital hybridization depends on two factors : hopping integral and orbital on - site energy . in the lsda+@xmath0 approach ,
@xmath0 does not change the hopping integrals , but shifts the on - site energies . as shown in fig .
[ fatband ] , the @xmath33 and @xmath34 bands are much closer in the lm solutions than those in the lsda solution , which is responsible for the enhancement of orbital hybridization . [ 0.65 ] orbital with majority spin , the right panels for the @xmath22 orbital with minority spin.,title="fig : " ] in a tight - binding model , the @xmath32 orbital on - site energy can be written as @xmath39 where @xmath40 stands for the on - site energy , @xmath41 for the orbital , @xmath35 for the spin , @xmath42 for the local magnetic moment , @xmath43 for the occupancy number .
@xmath44 is the stoner exchange integral , which is about 0.9 ev in metallic fe .
@xcite the minus ( plus ) sign of the term @xmath45 is for the spin that is parallel ( anti - parallel ) to the direction of the local magnetic moment .
@xmath46 corresponds to the double counting term of the lsda+@xmath0 method . in this work ,
we use the amf functional @xcite and have @xmath47 for the @xmath32 orbitals . for two orbitals with opposite spins ,
we get @xmath48 it is reasonable to assume @xmath49 , as the lsda solution shows that the @xmath29 and @xmath22 orbitals have similar occupancies .
then we can calculate the gap between the @xmath33 and @xmath34 orbitals . using the data in table [ occ ] ,
we get -1.58 ev for the lsda , -0.22 ev for the lm1 and 0.35 ev for the lm2 . these values are consistent with the character bands shown in fig .
[ fatband ] . to enhance the hybridization , the @xmath33 and @xmath34 orbitals need to be close .
[ diff ] shows some competitive factors for this requirement .
the term @xmath50 opens a gap between opposite spins , and therefore the @xmath0 term and the double counting term need to offset this splitting .
this gives a restriction on @xmath0 .
as shown in fig .
[ mom ] , there is no lm solution when @xmath0 is too small . for the lsda state , the exchange splitting
@xmath50 opens a gap of -1.58 ev .
if we start the lsda+@xmath0 calculation with the lsda solution , this gap is too large to be overcomed and the system is easier to converge to the hm state . to search for the lm solutions , we need to start the lsda+@xmath0 calculation with a small magnetic moment
this may be the technical reason why the lm states were not found before .
we have shown that the role of @xmath0 in the lm states is to enhance the orbital hybridization between antiferromagnetic neighbors by changing the on - site energies . here
we discuss the effect of @xmath32-@xmath32 interorbital hopping , which is the other factor influencing the strength of hybridization . from the structure of lafeaso
, we may realize that the direct interaction between the @xmath51 and @xmath52 orbitals is weak , and therefore the hybridization of @xmath33 and @xmath34 is via the as @xmath53 orbitals .
these @xmath53-@xmath32 interactions , such as @xmath54-@xmath51 and @xmath54-@xmath52 , are very sensitive to the fe - as - fe angle .
this characteristic can be used to tune the @xmath32-@xmath32 interorbital hoppings .
we adjust the fe - as - fe angle by changing the ratio between lattice constant @xmath55 and @xmath56 while keeping the volume constant .
our calculated magnetic moment of lafeaso as a function of @xmath55/@xmath56 is shown in fig .
the @xmath57-axis of the figure is labelled @xmath58 , where @xmath59 is the height of the as atom above the nearest fe layer . in the lm2 state ,
@xmath23 is anti - parallel to the direction of the total moment at small @xmath58 as shown in table [ occ ] . as @xmath58 increases
, @xmath23 becomes numerically larger and eventually the total moment gets the same direction as @xmath23 , which is indicated by the negative sign in fig .
the figure shows that the lsda state has a very weak dependence on the value of @xmath58 .
please note that the fe - as distance almost keeps a constant in this calculation . on the other hand ,
the lm states show very strong dependence on the feas@xmath3 geometry .
this phenomenon confirms the importance of the interorbital coupling to the lm states . we may conclude that the origin of the lm solutions is the coulomb - enhanced orbital hybridization between antiferromagnetic neighbors .
[ 0.6 ] with constant volume .
the magnetic moment for lm1 and lsda are shifted to fit the window .
the experimental values of @xmath58 for lafeaso , nafeas and lifeas are indicated by arrows .
the lsda+@xmath0 data are calculated with @xmath17=0 and @xmath0=3.26 ev . , title="fig : " ] fig .
1 indicates a critical value @xmath60 for the lsda+@xmath0 calculations : the lm states only exist at @xmath61 .
according to our explanation for the lm states , the coulomb interaction and feas@xmath3 structure are the two factors giving rise to the orbital - spin ordered states
. then @xmath60 may be reduced by changing the feas@xmath3 geometry .
we find that @xmath60 for nafeas is about 1.9 ev , which is much smaller than the value for lafeaso ( @xmath12.8 ev ) .
an interesting question is raised : can lsda or gga produce the lm states without @xmath0 ?
5 shows the low moment solution we find with gga ( without @xmath0 ) .
the gga - lm state appears at @xmath58=0.40 , and the total magnetic moment reaches a maximum at @xmath58=0.46 .
the gga - lm solution is not a simple low moment state , but an orbital - spin ordered state . at @xmath58=0.46 ,
the orbital moments are @xmath31=0.02 , @xmath27=0.04 , @xmath23=-0.23 , @xmath28=0.33 and @xmath26=0.01 @xmath2 .
we also present the band structure for @xmath58=0.46 in fig .
the very different band dispersion along @xmath57 direction and @xmath62 direction shows that this state is far away from the paramagnetic phase although it has a small total moment .
as may be seen , the gga - lm band structure is a little bit similar to the lm2 band structure shown in fig .
2 , while the main difference is the coulomb interaction nearly opens a gap at fermi level .
interestingly , we find that lsda can hardly produce this low moment state .
a possible reason is that lsda contains less electron correlation in the iron compounds than gga does .
the lsda+@xmath0 method is sometimes questioned due to the double counting problem . @xcite the choice of lsda+@xmath0 functional is somewhat arbitrary .
the gga - lm state proves that the orbital - spin ordering does not come from a specific lsda+@xmath0 functional , but originates in the tetrahedral structure of feas@xmath3 .
the gga - lm state provides a strong evidence for the orbital - spin ordering in fesc and also confirms our explanation for the lm states .
[ 0.61 ] ) .
the left panel is the magnetic moment as a function of @xmath63 with constant volume .
the right panel is the band structure for @xmath64 , which is corresponding to an orbital - spin ordered state with @xmath65 , @xmath66 and @xmath67 @xmath2 .
, title="fig : " ] it is known that the feas@xmath3 tetrahedron tends to have similar volume in different iron - arsenides , while its geometry is tuned by different inserted layers ( see table i of ref .
for example , lifeas has the largest as height @xmath59 in all iron - arsenides while its lattice constant @xmath56 is the smallest .
therefore , fig .
[ coa ] allows us to estimate the magnetic moments of the 111 materials .
the figure gives a moment of 0.1 @xmath2 for nafeas and -0.27 @xmath2 for lifeas .
here we consider the lm2 state because the lm2 state becomes more and more energetically favorable when @xmath58 is increasing as shown in fig .
[ coa ] . to check these estimated values
, we did a direct calculation using the experimental structures , which gives @xmath42=0.04 @xmath2 for nafeas and @xmath42=-0.20 @xmath2 for lifeas .
this agreement between the estimation and calculation shows that the feas@xmath3 geometry is a very important factor controlling the magnetic properties of fesc , which is responsible for the variety of magnetic moments in the 1111 and 111 materials .
our calculated magnetic moment for nafeas , @xmath42=0.04 @xmath2 , is in good agreement with the measured value .
@xcite but in contrast to the experiment , @xcite the ground state of lifeas in our calculation has the single - stripe afm order . actually , there might be experimental evidence for the striped afm order .
borisenko @xmath9 @xmath10 . measured the fermi surface of lifeas using angle - resolved photoemission .
@xcite they found that the pocket around the @xmath68 point is shaped like a butterfly , which indicates that the 4-fold rotation symmetry is broken .
this symmetry breaking might be due to the striped afm order .
4 may also explain the behavior of the parent compounds under pressure .
high - pressure experiment shows that the undoped lafeaso becomes superconducting at pressure @xmath69 gpa and reaches a maximum t@xmath4 of 21 k at @xmath70 gpa . for higher pressures the t@xmath4 decreases slowly .
@xcite lifeas shows very different pressure behavior from lafeaso .
lifeas is a superconductor without external pressure . under pressure , the t@xmath4 of lifeas decreases linearly at a rate of 1.5 k / gpa , @xcite which means that the maximal t@xmath4 is at zero pressure .
interestingly , a different high - pressure experiment indicated that the behavior of nafeas is more similar to lafeaso than lifeas , @xcite even though nafeas and lifeas belong to the same family .
to explain these experiments , we need to consider the structural distortion under pressure .
pressure experiments have shown that the fe - as bond length is very rigid while the fe - fe distance shrinks much faster when pressure increases , @xcite i.e. , @xmath58 increased by pressure .
it is found that chemical doping and pressure can both induce superconductivity in bafe@xmath7as@xmath7 and they have very similar influence on the feas@xmath3 geometry .
@xcite a noticeable fact is that chemical doping hardly changes the crystal volume .
@xcite this finding shows that the feas@xmath3 geometry is the key structural feature to superconductivity,@xcite whereas the volume reduction under pressure is less important .
therefore the @xmath57-axis direction of fig .
[ coa ] can be regarded as the direction of increasing pressure .
we consider the lm2 state in fig .
4 . for lafeaso and nafeas
, pressure first suppresses the magnetism .
after the magnetic moment reaches 0 , the pressure starts to enhance the magnetism .
it is known that the emergence of superconductivity in the fesc is related to the suppression of magnetism
. then the decreasing magnetism corresponds to the increasing t@xmath4 , conversely , the increasing magnetism corresponds to the decreasing t@xmath4 .
and the maximal t@xmath4 occurs at the magnetic moment @xmath42=0 . for lifeas
, pressure can only enhance the magnetism as shown in fig .
consequently , the t@xmath4 of lifeas has a negative pressure coefficient , @xmath71 .
therefore , these different high - pressure experiments can be consistently explained by our theory .
[ coa ] shows that the lm2 state is more and more energetically favorable than the lm1 state when @xmath58 increases ; hence the 122 compounds , which have the lm1 ground state at 0 pressure , will enter into the lm2 state when pressure increases .
then our explanation for the high - pressure experiments is also suitable for the 122 materials .
importantly , in this picture , the suppression of magnetism is because the orbital spin - moments cancel each other .
this suggests the superconducting phase is not a real nonmagnetic state but an orbital - spin ordered state . in this work
, we used the amf lsda+@xmath0 functional .
we also checked our calculation with another widely used scheme : the fully localized limit ( fll ) functional , @xcite which is suggested to be more suitable for strongly correlated systems .
we found one lm solution coexists with a hm solution in this functional .
the lm solution is very similar to the lm2 state and shows very similar @xmath58 dependence .
however , we did nt find the lm1 solution , and fll energetically favors the hm solution .
in summary , we have studied the magnetic properties of iron - arsenides using the lsda+@xmath0 approach .
two low moment states are found in our calculation , which reproduce the measured magnetic moments in the 1111 and 122 materials well .
the lm2 state shows good agreement with the low moment state found by previous tight - binding study .
our study reveals that the origin of the low magnetization is the strong orbital hybridization between antiferromagnetic neighbors . from the picture indicated above
we have learned : firstly , the geometry of feas@xmath3 is crucial to superconductivity and magnetism because it controls the @xmath32-@xmath32 interorbital coupling .
secondly , our theory suggests that the superconducting phase is not a conventional nonmagnetic state , but a state where the orbital spin - moments cancel each other .
this implies an inherent relationship between the magnetic correlation and superconductivity .
a. d. christianson , m. d. lumsden , o. delaire , m. b. stone , d. l. abernathy , m. a. mcguire , a. s. sefat , r. jin , b. c. sales , d. mandrus , e. d. mun , p. c. canfield , j. y. y. lin , m. lucas , m. kresch , j. b. keith , b. fultz , e. a. goremychkin , and r. j. mcqueeney , phys .
* 101 * , 157004 ( 2008 ) .
w. l. yang , a. p. sorini , c - c .
chen , b. moritz , w .- s .
lee , f. vernay , p. olalde - velasco , j. d. denlinger , b. delley , j .- h .
chu , j. g. analytis , i. r. fisher , z. a. ren , j. yang , w. lu , z. x. zhao , j. van den brink , z. hussain , z .- x .
shen , and t. p. devereaux , phys .
b * 80 , 014508 ( 2009 ) .
* s. a. j. kimber , a. kreyssig , y .- z .
zhang , h. o. jeschke , r. valenti , f. yokaichiya , e. colombier , j. yan , t. c. hansen , t. chatterji , r. j. mcqueeney , p. c. canfield , a. i. goldman and d. n. argyriou , nature mater .
* 8 , 471 - 475 ( 2009 ) . | the magnetic properties of iron - arsenides are investigated using the lsda+@xmath0 approach .
in addition to one high moment state , we find that a positive @xmath0 also produces two low moment states with [email protected] @xmath2 and [email protected] @xmath2 .
the electronic structures indicate that the low moment states originate in the strong orbital hybridization between antiferromagnetic fe neighbors .
therefore , the geometry of the feas@xmath3 tetrahedron is crucial to the low moment states , which is the key to understand the negative pressure coefficient of t@xmath4 in lifeas .
finally , our theory suggests that the superconducting phase is an orbital - spin ordered state , where the orbital spin - moments cancel each other . |
given that a combined configuration - interaction ( ci ) and many - body - perturbation - theory ( mbpt ) approach could describe well experimental data for energies and transition rates in neon , it was a natural continuation to investigate a more difficult , stronger core - correlated atom such as argon . in argon theories and experiments appear to be less accurate than in neon , with larger mutual disagreement ; moreover , heavier closed - shell atoms are even less studied and understood so that the argon atom can serve as an intermediate step in understanding many closed - shell atoms and ions .
although semiempirical calculations @xcite sometimes lead to a good agreement with experiment for some transition arrays and provide valuable information for the analysis of existing experimental data , they do not explain much the nature of correlations and their predictive power is limited to cases where ( a ) energies are well - known and ( b ) energy - level information is sufficient to infer other atomic properties such as transition probabilities .
in general two theories could give exactly the same energy levels without predicting the same transition probabilities .
for example , adding or not adding random - phase approximation ( rpa ) corrections to transition amplitudes will give two different answers for transition rates although wave functions obtained from the same effective hamiltonian are identical .
when rpa corrections were included in neon @xcite , the improvement of agreement with experiment was only marginal and it was not an issue , but as we will show in this paper this type correction is crucial in argon .
hence , theories , which do not take into account at least the rpa correction , can be misleading in argon and heavier noble gas atoms . still owing to cancellations of rpa contributions for the same transition array
, ratios obtained with a matrix method could be still accurate .
indeed , a fair agreement with experiment for low - lying argon transitions is obtained by @xcite even without considering rpa corrections .
however , the accuracy of ratios decreases if relativistic and correlation effects , that introduce asymmetry for the p@xmath0 and p@xmath1 hole rpa contributions to radial transition integrals , are significant .
another issue is the role of configuration interaction since many theories are limited by considering only a few configurations .
the convergence of ci has to be investigated .
furthermore , other many - body effects , such as an interaction of a valence electron with a core or a hole with a core or their combination , are quite significant in argon and heavier noble - gas atoms .
thus , the development of _ ab initio _ theory for heavy noble - gas atoms is needed to single out important correlation and relativistic effects .
some success is already achieved in this direction .
for example , elaborate _
ab initio _
calculations by @xcite performed for transitions from the lowest odd excited states @xmath2 _ { 1}$ ] and @xmath3
_ { 1}$ ] to the ground state give oscillator strengths in good agreement with average experimental values .
however , this theory was not extended to transitions between excited states .
therefore , one objective of this paper is to prove that transitions between excited states of argon can also be calculated with _
ab initio _ methods .
we also would like to show that for argon our simpler ci+mbpt theory has comparable accuracy to complicated single - double couple - cluster ( sdcc ) method of ref .
@xcite and thus can be of practical value for other theoretical groups . in this paper , first we will briefly describe our method of calculations ( more details are given in @xcite ) ; then , we will calculate energies and oscillator strengths of j=1 odd states and energies of j=2 even states and compare them with experiment .
this comparison gives an estimate on accuracy of our wave functions .
next , we will present line strengths and rates for transitions between excited states and compare our result with experiment and semiempirical calculations .
we have developed a convergent variant of ci+mbpt for particle - hole states of closed - shell atoms @xcite and successfully applied it in neon to calculate energies @xcite and oscillator strengths for transitions between excited states @xcite .
this perturbation theory is based on the first iteration of couple - cluster single - double equations @xcite .
all second - order mbpt terms @xcite are included , with some denominators being modified to take into account the strong interaction between a hole and a core electron or core and core electrons non - perurbatively @xcite .
the advantage of this approach compared to the iterated single - double method @xcite is simplicity and speed of calculations .
the accuracy of hole energies and fine - structure splittings is significantly improved already after adding second - order mbpt corrections as was illustrated previously in neon .
apart from coulomb correlation corrections , the breit magnetic interaction is also included , but small frequency - dependent breit , quantum - electrodynamic , reduced - mass , and mass - polarization corrections are omitted . to calculate particle - hole energies , we construct a model ci space , compute effective hamiltonian in this space , and
solve an eigenvalue problem .
along with energies we obtain wave functions , which are used to calculate transition amplitudes or other related quantities .
we investigate energies , oscillator strengths for transitions to the ground state , and line strengths or transition rates for transitions between excited states .
energies of neon particle - hole j=1 odd states and oscillator strengths were in very good agreement with experiment : pure _ ab - initio _ neon energies differed from experimental energies by 0.0069 a.u . , but after subtraction of the systematic shift ( which does not make much difference in transition calculations ) , the agreement was at the level of 0.0001 a.u . for almost all states . in this paper
we use the same approach to calculate energies and transition amplitudes of argon . reduced matrix elements for transitions between particle - hole excited states are given in ref .
@xcite : @xmath4 we found that second - order corrections to transition amplitudes or lowest order rpa corrections @xmath5 are significant in argon .
this type correction requires careful analysis since the possible interaction between an observing " hole state @xmath6 and core states @xmath7 over which summation is carried out should be taken into account in all orders , at least a bulk part of this interaction .
similar to our approach for energies , we include the interaction in all orders by adding the extra term @xmath8 in the denominators of rpa expressions .
this procedure reduces rpa corrections roughly twice .
if we note that the value of rpa correction is comparable to the values of lowest order matrix elements , the modification of the denominators is important and brings our theory to a good agreement with experiment which will be illustrated later in this paper .
table [ arg21 ] shows our calculated energies for j=1 odd states .
fine structure for the 4s , 5s , or 4f groups of states is well reproduced since the deviation from experiment for each fine structure group of levels is similar .
the last column shows shifted deviations from experiment , which demonstrate that there is a systematic shift for energy levels in our calculations .
this shift is due to the inaccuracy of our calculations of hole energies .
the shifted deviations illustrate a good precision for higher excited states which have smaller correlation effects .
we also show the convergence with the number of particle - hole configurations .
the change in energy is significant when we increase the size of ci space from 22 to 32 and to 52 ; however , from 52 to 62 the change is small , and we consider that ci-52 case is optimal .
.energies of argon odd j=1 states . for unique specification of levels ,
accurate nist energies are provided .
the sizes of ci model spaces are denoted ci-22 , etc .
@xmath9 is the deviation from experiment for the 7th level .
all energies are expressed in atomic units [ cols="^,^,^,^,^,^,^",options="header " , ] table [ argls ] shows the comparison of our calculations with experiment for line strengths between j=1 odd and j=2 even state transitions .
very large disagreement is observed when no rpa correction is added .
however , if conventional rpa correction ( eq.[rpa1 ] ) is included , the difference is also large , though of an opposite sign . and
only if we include the appropriate , modified - denominator rpa correction ( mrpa ) , which constitutes roughly a half of the conventional rpa correction , do our values agree with experiment at the level of the experimental accuracy . the experimental accuracy in table [ argls ]
is denoted by letters in accordance with the nist convention : for example , the accuracy of class `` c '' is about 10% . still in the case of the p@xmath104s - p@xmath104p transition
the disagreement is quite large and independent of rpa corrections which might indicate that the problem is in the experiment .
if we exclude this case and the inaccurate measurement of class `` d '' , then mrpa calculations can be considered accurate at the level of 10 - 20% .
we also compare our final results ( mrpa calculations ) with some available semiempirical calculations by @xcite . in some cases , semiempirical results are close to experiment , but in others they can deviate substantially , by orders of magnitude , and they also disagree with each other .
for example , for weak transitions , transition rates of @xcite are about 10 times larger than experimental values ; for these transitions results of @xcite are much better .
our theory agrees with experiment more consistently in cases where experiments have good precision .
for one particular transition at 430 nm , many measurements have been conducted , and this transition can be considered to be known most accurately among other transitions between excited states of argon . in fig .
[ arg430 ] we compare our theoretical value with many available experimental values .
while six experiments support nist quoted value , a greater number of experiments support our theoretical value .
therefore , we tend to concluded that nist values can be somewhat wrong .
we also find that semiempirical calculations are not able to predict these particular experiments ; namely , most extensive semiempirical calculations @xcite give the transition rate @xmath11 s@xmath12 that disagrees with all experiments . returning to our table [ argls ] , we can argue that some large deviations of our theory from experiment can be caused by inaccuracy of quoted experimental errors . for better tests of our theory ,
more accurate experiments are needed .
for this reason , we have measured the ratio of line strengths using an accurate laser absorption technique @xcite .
our experimental ratio , @xmath13 , for the two transitions at 9787 and 9227 completely agrees with our theoretical value , 3.22 .
it would be extremely interesting if this experimental technique was further developed and applied to other transitions in argon or in other noble gas atoms .
for example , experiments in neon using visible lasers could test the accuracy of this experimental method , since our theory is expected to be more accurate in neon than in argon .
in this paper , we have applied ci+mbpt theory to calculate energies and transition properties of argon particle - hole states .
our theoretical energies and transition rates are in good agreement with most experiments .
we found that random - phase approximation corrections are essential .
we also analyzed dependence of accuracy on the size of configuration space and determined the optimal size of ci matrix .
our ci+mbpt program can be used for calculations of many other transitions of closed - shell atoms and ions . | argon energies of j=1 odd and j=2 even states , oscillator strengths of the two lowest j=1 odd states , and line strengths for transitions between j=1 odd and j=2 even excited states are calculated with the ci+mbpt method .
the results are in good agreement with experiment . |
the topological insulators are characterized by spin polarized helical boundary states protected by time reversal symmetry.@xcite in three dimensions , the bi@xmath0se@xmath1 material has attracted great attention because of its well defined dirac cone like surface state dispersion.@xcite though perfect two - dimensional surface states account for the most essential properties of topological insulator , real samples also have lateral pairs of surfaces . for strong topological insulators
, each surface carries an odd number of gapless topological surface modes.@xcite when the bulk material is constricted in two directions , we get a ribbon ( or , nano - wire ) of topological insulator.@xcite the aharonov - bohm effect of the bi@xmath2se@xmath3 ribbon showed interesting @xmath4 period oscillation and a conductance maximum at zero flux@xcite , which are in contradiction with both diffusive and ballistic transport and are explained theoretically in terms of impurity scattering effects.@xcite more conventionally , the ribbon ( nano - wire ) of bi@xmath2te@xmath3 , which is also a topological insulator and described by a model similar to that of bi@xmath0se@xmath3 @xcite , is considered as a promising thermoelectrical material for applications.@xcite former theoretical studies of topological insulator ribbon usually start from a cylindrical sample.@xcite while an axi - symmetric configuration is more convenient for analytical analysis , a real topological ribbon usually has a rectangular cross - section.@xcite we thus study the charge and spin distributions of the surface modes of such a realistic topological insulator ribbon . for a @xmath5 ( or equivalently , @xmath6 ) type of effective model for the surface states ,
the charge current operator is proportional to the spin operator .
the optical conductivity for the surface states , which is defined in terms of the current - current correlation function , is correspondingly directly related to the dynamical spin susceptibility.@xcite but for a topological insulator ribbon , in which two pairs of lateral surfaces are present , the relationship between the optical conductivity and the dynamical spin susceptibility would in general be more complicated .
however , the results always depend uniquely on the characters of the model used , which also determine the properties of the surface states .
the study of these responses may thus give some criteria to discriminate among available models for the topological insulator bi@xmath0se@xmath1 .
in a previous work , two different models used in former works for bi@xmath0se@xmath1 were pointed out.@xcite here , we propose that the more adequate model for the material could be identified by measuring the dynamical spin susceptibility of a bi@xmath0se@xmath1 ribbon . while optical conductivities are identical , the dynamical spin susceptibilities show clear qualitative differences between results obtained from the two models .
the results are explained in terms of selection rule analysis based on direct numerical calculations . for optical conductivity ,
qualitatively different selection rules are in effect between transitions along the ribbon axis ( @xmath7 direction in this work ) and transitions along other directions .
since we are interested mainly in the dynamical responses related to the topological surface states , it is reasonable to start from the continuum model for a topological insulator describing low energy states close to the @xmath8 point of the brillouin zone ( bz).@xcite formerly , we recognized two different kinds of models that exist in the literature for bulk bi@xmath0x@xmath1 ( x is se or te ) materials.@xcite close to the @xmath8 point , the two models could both be written compactly in terms of the dirac matrices as@xcite @xmath9 every unit cell contains two spin and two orbital degrees of freedom . the model is hence written in terms of 4 by 4 matrices .
the two orbitals concentrate mainly on the top and bottom ( looking along the @xmath10 direction ) se layer of the various bi@xmath0se@xmath1 quintuple units , and are labeled as 1 and 2.@xcite the basis is taken as @xmath11=[@xmath12 , @xmath13 , @xmath14 , @xmath15@xmath16 , for a certain wave vector in the three dimensional bz . since the first term proportional to the unit matrix is nonessential to topological properties of the system , it is ignored in the following analysis .
the remaining model is particle hole symmetric .
close to the @xmath8 point , the four remaining coefficient functions are @xmath17=@xmath18 , in which @xmath19 , @xmath20 and @xmath21 ( @xmath22 is half of the bulk band gap).@xcite in terms of pauli matrices @xmath23
( @xmath24=0 , @xmath25 , 3 ) in the spin subspace and @xmath26 ( @xmath24=0 , @xmath25 , 3 ) in the orbital subspace , the first three dirac matrices are defined as@xcite @xmath27=@xmath28 , @xmath29=@xmath30 , @xmath31=@xmath32 .
two choices of the last dirac matrix @xmath33 define the two different models in literature , which are : ( i)@xmath34@xcite and ( ii)@xmath35@xcite . though a full gap exists in the bulk , gapless modes reside on the surface . for surfaces directed along different directions ,
the surface states are described by different effective models .
take for example the surface states for a sample occupying the lower half @xmath36 space , the effective model for it was solved to be@xcite @xmath37 for model i. the charge current matrix is thus @xmath38 . where the pauli matrices @xmath39 and @xmath40 are in terms of the two basis @xmath41=@xmath42 and @xmath43=@xmath44 .
@xmath45 is one for @xmath46=@xmath47 and zero otherwise . while for model ii , the effective model for the surface states of the same system is@xcite @xmath48 the corresponding charge current is @xmath49=@xmath50
. the two bases are @xmath51=@xmath52^{t}$ ] and @xmath53=@xmath54^{t}$ ] . for the surface states introduced above residing on the @xmath55 surface , the two bases have definite spin characters for both model i and model ii .
now consider the surface states of a sample occupying the half space @xmath56 .
similar to surface states on the @xmath57 surface , the possible zero energy surface states on @xmath58 are obtained by solving a set of four coupled second order differential equations @xmath59 for @xmath60 together with the open boundary conditions @xmath61=@xmath62=0.@xcite for model i , the basis for the surface states are obtained as @xmath51=@xmath63^{t}$ ] and @xmath53=@xmath64^{t}$ ] by solving the above coupled differential equations .
now , both orbitals contribute to the surface states .
the corresponding effective model turns out to be @xmath65 while for model ii , the surface states are described by an effective model as @xmath66 with the same basis set as for model i. the two basis states now do not have definite spin characters . on the other hand ,
the dirac cone of the surface states becomes anisotropic , which is a manifestation of the uniaxial anisotropy of bulk states between the @xmath55 plane and the @xmath36 direction .
@xcite the above effective models for topological surface states describe infinite surfaces .
a real sample is however finite in all three directions . for a ribbon geometry realized and studied recently , the sample could be considered as infinite along the direction of the ribbon axis ( taken as @xmath7 ) while finite in the other two directions ( taken as @xmath67 and @xmath36).@xcite then , while @xmath68 could still be taken as a good quantum number , real space viewpoint should be adopted for @xmath67 and @xmath36 dependent quantities . for the sake of simplicity , we discretize the @xmath67 and @xmath36 coordinates into square lattices .
the model is thus written as @xmath69t_{z } \psi_{n_{y},n_{z}+1}(k_{x } ) + h.c.\ } \notag \\
+ \sum\limits_{n_y , n_z , k_x}\{\psi^{\dagger}_{n_{y}n_{z}}(k_{x})[-\gamma_{0}-i\gamma_{2}]\frac{3}{2}t \psi_{n_{y}+1,n_{z}}(k_{x } ) + h.c.\},\end{aligned}\ ] ] where @xmath70=@xmath71 .
@xmath72 and @xmath73 label the unit cells along the @xmath67 and @xmath36 directions .
the lattice constants along the three directions are taken as length units .
the paramagnetic current operators along three directions could be obtained by the continuity equations.@xcite they are written as @xmath74 @xmath75,\ ] ] @xmath76,\ ] ] where @xmath77=@xmath78 and @xmath79=@xmath80 . in this work
, we would focus on the zero temperature response behaviors of the system .
the dynamical conductivity is obtained from the retarded current - current correlation functions by the kubo s formula.@xcite ignoring contribution from the diamagnetic current ( which contributes to the zero frequency drude weight ) , the expression is @xmath81 where the retarded current - current correlation function is defined as @xmath82|0\rangle \notag \\
= \frac{1}{v}\sum\limits_{n}[\frac{\langle0|j^{p}_{\alpha}|n\rangle\langle n|j^{p}_{\beta}|0\rangle } { \omega+e_{0}-e_{n}+i\eta}-\frac{\langle0|j^{p}_{\beta}|n\rangle\langle n|j^{p}_{\alpha}|0\rangle } { \omega+e_{n}-e_{0}+i\eta}]\end{aligned}\ ] ] @xmath83 is a complete set of basis states , with @xmath84 denoting the ground state .
@xmath85 is energy of the state @xmath86 .
@xmath87 is the positive infinitesimal and taken as a small positive number in realistic calculations .
the spin operator of the system is defined as @xmath88 = @xmath89 .
the zero temperature dynamical spin susceptibility ( for @xmath90 ) is thus defined as @xmath91|0\rangle \notag \\
= \frac{1}{v}\sum\limits_{n}[\frac{\langle0|s_{i}|n\rangle\langle n|s_{j}|0\rangle } { \omega+e_{0}-e_{n}+i\eta}-\frac{\langle0|s_{j}|n\rangle\langle n|s_{i}|0\rangle } { \omega+e_{n}-e_{0}+i\eta}].\end{aligned}\ ] ]
without loss of generality , we consider a ribbon with @xmath92=@xmath93=20 and for @xmath94=@xmath95=0.5 , @xmath96=-0.7 .
1(a ) is the dispersion of the ribbon . within the bulk gap region ( @xmath97 ) ,
sub - bands form with a finite minimal gap as a result of discretization of the topological surface states induced by lateral constriction . as shown in fig .
1(a ) , the sub - bands are labeled symmetrically with respect to @xmath60 as @xmath98@xmath99 , @xmath100 , @xmath101@xmath102 , with positive ( negative ) sign denoting positive ( negative ) energy sub - bands and the larger numbering representing a larger distance to @xmath60 .
every state is twofold degenerate due to combined time reversal symmetry and inversion symmetry of the material
. = @xmath93=20 .
( b ) charge distribution ( multiplied by @xmath103 ) of the surface modes @xmath104=3 for @xmath68=0 , results for other @xmath104 ( for @xmath68=0 ) within the bulk gap region are similar .
( c ) and ( d ) are the charge distribution ( multiplied by @xmath103 ) of the two fold degenerate @xmath104=1 surface eigenmode for @xmath68=0.1@xmath105.,width=340,height=302 ] the charge and spin distributions of the topological surface states are essential to its physical properties . in the ribbon geometry ,
an interesting feature is related to the hybridization and redistribution of different surface states .
we give the spin and charge distributions for several typical surface modes on the sub - bands within the bulk gap . as shown in fig .
1(b ) for the charge distribution of a typical surface mode for @xmath68=0 , the charge is distributed centro - symmetrically close to the boundary of the ribbon s cross - section . around the cross - section ,
more charge is distributed close to the four corners which is easily understood from the smaller effective radius there compared to the flat part of the sample . as shown in figs .
1(c ) and 1(d ) , the charge distributions for the twofold degenerate modes of @xmath104=1 for @xmath68=0.1@xmath105 are asymmetric and centered around one corner of the ribbon , displaying clearly the effect of hybridization of two kinds of surface modes across the corner .
the three components of the spin expectation value on every site are presented in ( a ) , ( b ) and ( c ) ( d , e and f ) for model i ( model ii ) .
the results in ( a ) , ( b ) and ( c ) ( in d , e and f ) are multiplied by @xmath103 ( @xmath106).,width=340,height=377 ] while the charge distributions are qualitatively identical along the @xmath67 and @xmath36 directions and are the same for the two different models , the spin distribution is more complicated . for a state residing on an infinite @xmath55 surface ,
the spin only has in - plane component . while for states on an infinite @xmath107 surface ,
the basis of the surface states indicate that the spin has both in plane and out of plane components . in the present ribbon geometry
, the surface modes resulting from hybridization of the two types of surface modes would be different from an ideal surface .
2 shows distributions of the three spin components for a typical surface mode , which is one of the two degenerate modes for @xmath104=1 and @xmath68=0 . figs .
2(a , b , c ) are for model i. a peculiar feature is that the expectation value of the @xmath67 spin component is everywhere zero . however , spin distributions of the corresponding surface state for model ii shown in figs .
2(d , e , f ) have all the three spin components nonzero along the edge of the cross - section . , @xmath67 and @xmath36 directions for three typical chemical potentials ( a ) @xmath108 , ( b ) @xmath109 , and ( c ) @xmath110 .
results for the two models are identical .
@xmath111 is the bulk band gap .
@xmath87=0.01.,width=264,height=453 ] having the above charge and spin distributions of the surface sub - bands in mind , it is natural to expect that the charge response of the two models would be similar , but the spin responses would possibly show very different behaviors .
3 shows the optical conductivity ( in unit of @xmath112 ) of a ribbon with @xmath93=20 and @xmath92=20 .
the same results are found for the two models .
the optical conductivity is defined as real part of the longitudinal dynamical conductivity and is related to optical absorption .
one interesting feature is the low energy peaks of optical absorptions related to optical transitions between the surface sub - bands inside of the bulk band gap for @xmath113.@xcite when the chemical potential lies at the charge neutral point , only the @xmath7 component of the optical conductivity is significant .
the @xmath67 and @xmath36 components of the optical conductivity are smaller and nearly featureless . when the chemical potential is shifted away from the dirac point by @xmath114 , the low frequency part below 2@xmath115 for @xmath116 is depleted while the higher frequency part remains unchanged . in contrast
, some low frequency ( @xmath117@xmath118@xmath119 ) peaks emerges for @xmath120 and @xmath121 .
this reflects a difference in the selection rules of optical transitions along the three directions .
since @xmath68 is still a good quantum number , only vertical optical transitions between states with the same @xmath68 are allowed along all three directions , as could be seen from the current operators defined in eq .
differences between @xmath116 and optical conductivities along the other two directions arise from selection rules with respect to @xmath104 in the optical transitions .
numerically , only transitions between pairs of sub - bands labeled by @xmath104 and @xmath122 are allowed for optical transitions along the @xmath7 direction .
while for optical transitions along @xmath67 and @xmath36 , the transitions involving sub - band @xmath104 is largest with respect to @xmath123 .
so , while for @xmath116 a series of peaks below 2@xmath22 are expected corresponding to transitions between @xmath104 and @xmath122 sub - bands , @xmath120 and @xmath121 only have low frequency peaks corresponding to transitions between adjacent sub - bands . for nonzero @xmath114 , the low frequency ( @xmath124 ) optical transitions contributing to @xmath116
are strictly forbidden since the involved sub - band pairs are all occupied .
thus the corresponding low frequency part of @xmath116 is depleted .
but for @xmath120 and @xmath121 , the number of states that could contribute to low frequency optical absorptions continuously increases , resulting in a monotonic enhancement of the low frequency ( @xmath124 ) peaks with increasing @xmath115 .
this nicely explains the evolution of the low frequency optical conductivities in fig .
3 . we now proceed to calculate the dynamical spin susceptibilities . in accordance with the optical conductivity , we focus on the imaginary part of the dynamical spin susceptibilities . only the longitudinal components would be analyzed here . for an ideal @xmath55 surface ,
the optical conductivity is identical to the dynamical spin susceptibility . in the presence of @xmath107 surface , another correspondence between optical conductivity and spin susceptibility specific to the @xmath107 surface states mixes in . in fig .
4 we show the imaginary parts of the longitudinal dynamical spin susceptibilities normalized by the frequency . in agreement with expectation based on the different spin distributions of the surface sub - bands ,
the dynamical spin responses for the two different models show qualitatively different behaviors and thus could be used as a means to tell the right model for bi@xmath2se@xmath3 . for model i , ( b ) @xmath109 for model i , ( c ) @xmath108 for model ii , ( d ) @xmath109 for model ii .
@xmath111 is the bulk band gap .
@xmath87=0.01.,width=340,height=302 ] though the two types of surfaces are coupled together , some qualitative correspondences between the optical conductivity and the dynamical spin susceptibility could still be identified.@xcite in particular , for an infinite @xmath55 surface , the @xmath7 component of the optical conductivity is proportional to the @xmath7 component of the dynamical spin susceptibility for model i while it is proportional to the @xmath67 component of the dynamical spin susceptibility for model ii .
qualitatively similar correspondence could be identified from comparing fig .
4 with fig .
the @xmath67 and @xmath36 components of the optical conductivity suffer stronger influences from the lateral constriction and the induced surface states hybridization , the correspondences to the dynamical spin susceptibilities are also not very straightforward . note that , despite the similarity with the @xmath7 component of the optical conductivity , a subtle difference is that the first peak of the dynamical spin susceptibility occurs at a frequency which is twice that of the first peak in @xmath125 .
this reflects a general difference between selection rules for electronic dipolar transitions ( contributing to @xmath126 ) and magnetic dipolar transitions ( contributing to @xmath127 ) , which coincides only very rarely for a state such as the surface state on the @xmath55 surface .
explicitly , the matrix elements for the spin operators are nonzero only for transitions connecting second neighbor sub - bands , that is between @xmath104 with @xmath128 , profoundly different from the above - mentioned selection rules for the optical conductivity .
we would like to point out an intricate point on identifying the right microscopic model for bi@xmath2se@xmath3 .
experimentally , the spin polarization of the @xmath55 surface states was found to be perpendicular to the 2d wave vector.@xcite though model ii gives naturally the right spin polarization , it could not directly be chosen as the right model . changing the spin - orbital coupling in the @xmath55 plane from @xmath6 to @xmath5 in the bulk model , the same spin polarization as observed in experiment
is realized also for model i.@xcite since this substitution amounts to a simple redefinition of the spin axes in the @xmath55 plane , @xmath129 and @xmath130 simply exchange with each other and thus similar distinction between the two models by the dynamical spin susceptibilities is still feasible .
finally , we mention the experimental measurements of the dynamical spin susceptibilities
. the electron magnetic resonance ( emr ) could approach @xmath1313 thz which amounts to approximately 12 mev.@xcite for a ribbon with much larger cross - section than considered above but similar to experiment,@xcite the sub - bands inside of the bulk band gap would be much denser than for @xmath132 .
the corresponding low frequency peaks in the dynamical spin susceptibility would also within the reach of emr.@xcite another technique is the inelastic neutron scattering ( ins ) .
though it is usually hard to see surface effects in terms of ins , when the chemical potential is tuned close to the dirac point inside of the bulk band gap , the surface would possibly give the dominant signal .
at last , the spin flip surface raman scattering could be another possible way to measure the dynamical spin susceptibility.@xcite
we have calculated the charge and spin distributions and the corresponding dynamical spin and charge responses of a topological insulator ribbon .
two models giving different surface states are analyzed as a comparison .
constriction along two lateral directions change the gapless dirac cone like surface states into sub - bands inside of the bulk band gap .
the corresponding charge distributions of the sub - band states resulting from hybridization of four different surface states are identical for the two models , giving rise to identical optical conductivities for the two different models .
the spin distributions of the surface states are however quite different between the two models .
the dynamical spin susceptibilities are thus also quite different and could be used to identify the right model for bi@xmath0se@xmath1 .
d. hsieh , y. xia , d. qian , l. wray , j. h. dil , f. meier , j. osterwalder , l. patthey , j. g. checkelsky , n. p. ong , a. v. fedorov , h. lin , a. bansil , d. grauer , y. s. hor , r. j. cava , and m. z. hasan , nature ( london ) * 460 * , 1101 ( 2009 ) . | we study theoretically the distributions of charge and spin polarization of a topological insulator ribbon , with a realistic rectangular cross section . due to constriction in two lateral directions ,
the surface states discretize into a series of subbands inside of the bulk band gap .
the charge and spin distribution show interesting characters which are different from an ideal topological surface state .
the effect of merging of four different surface states into the new one as an entity are analyzed .
optical conductivity and dynamical spin susceptibility of the ribbon are studied .
different from a single ideal surface , the optical response and dynamical spin susceptibility of a ribbon do not have a clear correspondence .
the dynamical spin susceptibility could be used to identify the more adequate model for bi@xmath0se@xmath1 . |
the electronic structure of the carbon nanoparticles is of great interest in the todays physical research .
these nanostructures have a large potential use as electronic nanodevices in the computer science .
the main structure of this kind is the plain graphene , but other kinds of the nanostructures like fullerene , nanotubes etc .
were already largely investigated .
the main characteristics of the electronic properties is the local density of states ( @xmath0 ) .
it can be calculated using the different methods ; here we use the continuum gauge field - theory approach @xcite which leads to the solution of a 2-component dirac - like equation .
the result we get is the sum of the squares of the components of the normalized solution .
this method was used in @xcite for the calculation of the electronic structure in the close surrounding of different kinds of the defects and of fullerene , nanocone , nanotubes and wormhole .
one phenomenon has not been yet considered enough in the graphene nanostructures : the effect of the spin - orbit coupling ( @xmath1 ) .
the reason is that this effect is very weak in the case of the plain graphene , but it can grow considerably in other kinds of the carbon nanostructures because of the curvature and the orbital overlap of the next - nearest neighbouring atomic sites . furthermore , the influence of the impurities could play a crucial rule .
then , we distinguish the intrinsic and the extrinsic ( rashba ) term of the @xmath1 .
a lot of theoretical work concerning this effect was performed in @xcite .
there are more ways how to incorporate these terms into the hamiltonian for the ground state of the given nanostructure . for the case of the graphitic nanocone @xcite ,
different methods are described in @xcite and @xcite : in these papers , the terms corresponding to the @xmath1 are added into different blocks of the corresponding matrix and , consequently , it is mixed with different terms of the remaining part of the hamiltonian . here
, we will be more interested in the method used in @xcite , where the used formalism is an analog of the procedure used for the case of the nanotubes @xcite . in @xcite ,
the boundary effects were not considered .
their significance is ( among others ) connected with the needs of the quadratic integrability of the resulting wave function .
they come from the rapid change of the geometry close to the conical tip and from the finite size of the conical nanostructure .
the change of the geometry could give rise to the mass of the fermion @xcite , but it could be also simulated by an addition of a fictious charge into the conical tip @xcite .
next , the influence of the pseudopotential is also significant @xcite .
finding of the analytical solution of the resulting hamiltonians is a very difficult task . in most of our calculations ,
we use the numerical methods . to achieve a higher precision , the green function approach
is also used . in this paper
, we review the results of the calculations concerning the influence of the @xmath1 on the electronic structure of the graphitic nanocone in @xcite , discuss the influence of the boundary effects and incorporate the green function formalism . in section
ii , we remind the main results from @xcite . in section iii
, we present the influence of the possible boundary effects connected with the topology of the molecular surface : the relativistic acquisition of the mass connected with the curvature and the coulomb interaction .
finally , in section iv , we derive the green function perturbation scheme to the second order for the purpose of getting more precise results of the calculation of the @xmath0 .
in @xcite , we derived the hamiltonian for the curvature induced @xmath1 in the graphitic nanocone using the method of ando @xcite .
we started with the case of the hamiltonian without the @xmath1 @xcite and after incorporating the appropriate @xmath1 terms , we finally got @xmath2 where all the parameters connected with @xmath1 and with the geometry of the nanocone are described in the mentioned paper . to calculate the @xmath0 , we solve the equation @xmath3 and
finally get the matrix equation ( 9 ) in @xcite . there , this equation is solved in a numerical way and the resulting @xmath0 for @xmath4 is sketched in fig . 3 of that paper .
the numerical method of finding the solution used in @xcite can be used for other modes . in fig .
[ ldos3dspin ] , we see the @xmath0 for the same numbers of the defects , but here it corresponds to the sum of the solutions corresponding to @xmath5 .
we can see that while for the case of 1 and 2 defects in the tip , the @xmath0 should grow to infinity for @xmath6 for an arbitrary value of energy , this effect is restricted only to zero energy in the case of 3 defects in the tip . for the graphitic nanocone influenced by the @xmath1 . here
, the @xmath0 corresponds to the sum of the solutions corresponding to @xmath5 .
the number of the defects in the tip in the particular cases : @xmath7 ( left ) , @xmath8 ( middle ) and @xmath9 ( right).,width=604 ] from these results follows that there could be a strong localization of the electrons in the tip .
but we did not consider different boundary effects which could significantly influence this behavior .
our effort is to find a quadratically integrable solution which preserves the main features of the found approach .
this means that for 1 and 2 defects the rise of the @xmath0 close to @xmath6 will be suppressed or restricted to a much smaller area around @xmath6 , and the peak for @xmath6 and @xmath10 in the case of 3 defects will be preserved . in the last case of 3 defects ,
the calculations show that the main contribution to the revealed peak is the mode @xmath11 and that considering the next modes the resulting character of the @xmath0 for @xmath6 will approach the behavior of the case of 1 and 2 defects ( fig .
[ ldos3dspin3d ] ) . for the graphitic nanocone with 3 defects in the tip influenced by the @xmath1 : @xmath11 ( left ) and the case corresponding to the sum with @xmath12 ( right).,width=604 ]
the hamiltonian ( [ hamsoc ] ) is not the only possibility how to express the hamiltonian for the @xmath1 : we can also use the form of the hamiltonian where the spin orbit coefficients and the differential operators occupy different blocks of the corresponding matrix and find the analytical solution .
this possibility is outlined in @xcite , where the rashba term is excluded and the intrinsic terms are moved to the diagonal positions of the hamiltonian .
so the resulting hamiltonian ( eq . ( 10 ) in @xcite ) has the form @xmath13 with @xmath14 and @xmath15 .
let us stress that here the intrinsic term does not depend on @xmath16 and the @xmath1 depends on the vortex angle only .
the solution for this hamiltonian is given by the modified bessel function of the second kind .
on the contrary , the analytical solution of ( [ hamsoc ] ) was not found ; we can only say that the found numerical solution @xcite is similar to the bessel function of the first kind . working on the hamiltonian ( [ hamsoc ] ) investigated in @xcite ,
let us investigate the boundary effects to get more precise information about the electronic structure close to @xmath6 .
in the real graphitic nanocone , the tip has not a sharp form as one would expect for the cone geometry , but the geometry becomes smooth in this region ( fig .
[ geom ] ) ; it means , there is such a value of @xmath17 that our predictions fail for @xmath18 . from the possible ways how to remove this discrepancy
, we outline these 2 methods : having to find a way to simulate the smooth geometry in the conical tip , we either suppose that due to a big curvature and the connected relativistic effects , the massless fermion , for which we solve equation ( [ direq ] ) , acquires a non - negligible mass @xmath19 @xcite or a charge is considered in the conical tip from which the coulomb interaction comes influencing the results . for @xmath18 , we can write the corrected version of the matrix equation ( 9 ) from @xcite : @xmath20 for higher values of @xmath16 , where the curvature is considerably smaller , the mass is considered to be zero and the form of the matrix equation is changed into the mentioned form .
we can solve this equation numerically and analytically , as well . in the case of the numerical solution ,
especially the case of 3 defects would cause a shift and decoupling of the peaks in the corresponding graph .
however , the main purpose of using the boundary conditions - getting the solution which is quadratically integrable - is not achieved here . for the purpose of getting the analytical solution
, we suppose in ( [ syst1 ] ) that the mass @xmath19 acquired by the fermion is considerably larger than other effects contained in the hamiltonian , i.e. @xmath21 and @xmath22 ( and the same up to the 4-th order of the powers and the differentiations ) .
then , the solution can be very roughly approximated as @xmath23 @xmath24 @xmath25 @xmath26.\ ] ] the graphs of the @xmath0 corresponding to this solution coincide with the graphs for the numerical solution with @xmath27 , so the potential peaks for an arbitrary number of the defects disappear here .
+ the influence of the charge considered in the conical tip is expressed in the hamiltonian by the presence of the diagonal term @xmath28 , where @xmath29 is the fine structure constant .
so this term substitutes the mass term in ( [ syst1 ] ) : @xmath30 here , the coefficients @xmath31 can be considered nonzero , as well as zero , depending on whether we consider the parallel influence of both the coulomb interaction and the @xmath1 or the coulomb interaction only . the analytical solution for the second case
is given in @xcite and we outline the corresponding @xmath0 in fig .
[ graph - coulomb - bezs ] .
( it is possible to consider nonzero mass in this case but here we suppose , in agreement with ( [ syst2 ] ) , @xmath32 ) . in the expression for the analytical solution ,
@xmath33 has half - integer values , here we considered @xmath34 .
the outlined solution , similarly as the solution for the case of the @xmath1 only in fig .
( [ ldos3dspin ] ) , signalizes the localization of the electrons for @xmath6 , but on the contrary to that figure , for higher @xmath16 and energies close to zero , the @xmath0 decreases to zero . for the first case
, we use the analog of the numerical method used in @xcite : we express the solution in the form of an infinite series ( eqs .
( a1 ) , ( a2 ) in @xcite ) and substitute it into the system ( [ syst2 ] ) . by the addition of the requirement @xmath35
, we get a system of recurrence equations for the coefficients @xmath36 . from the boundary conditions
follows @xmath37 and the same for the coefficients @xmath38 .
then , the value of the parameter @xmath39 is determined from the requirement of the nonzero solution of the system for the coefficients @xmath40 , i. e. the determinant of the corresponding matrix should be nonzero .
in this way , we get @xmath41 we have to choose one of the 4 possible values of @xmath39 ; here we chose the value with plus signs on the critical places .
then , the coefficients @xmath40 create the components of the corresponding eigenvector .
now , for @xmath42 , we can easily calculate the coefficients @xmath43 in the chosen limit .
the coefficient @xmath44 can be chosen arbitrarily ; we chose @xmath45 .
the resulting graphs of the @xmath0 are in fig .
[ graph - coulomb ] . here , similarly as in fig .
[ ldos3dspin ] , @xmath46 . comparing figs .
[ graph - coulomb - bezs ] and [ graph - coulomb ] we see that the same problem appears in both cases : the electrons are localized in the tip for an arbitrary energy for an arbitrary number of the defects and the uniqueness of the peak for zero energy , which corresponded to the mode @xmath11 and 3 defects in the case of the @xmath1 only ( fig . [ ldos3dspin ] ) , is distorted by the divergence of the @xmath0 at all energies for @xmath6 .
+ for the graphitic nanocone influenced by the coulomb interaction ( with the exclusion of the @xmath1 ) for different distances @xmath16 from the tip , @xmath34 and for 1 , 2 and 3 defects .
we see that similarly as in the case of the @xmath1 only , we get the localization of the electrons in the tip for 1 and 2 defects at all energies but , furthermore , there is no energy restriction for the case of 3 defects as well.,width=642 ] for the graphitic nanocone influenced by the coulomb interaction ( including the influence of the @xmath1 ) for different distances @xmath16 from the tip , @xmath46 and for different numbers of the defects.,width=642 ] let us investigate the character of the divergence of the @xmath0 for @xmath6 closer . in fig .
[ ldosrcs ] , we see the dependence of the @xmath0 on @xmath16 variable for zero energy in the case of the influence of the @xmath1 only and of the simultaneous influence of the @xmath1 and the coulomb interaction . we see here that in comparison with the first case , in the second case the decrease of the @xmath0 close to @xmath6 is much faster and one could suppose that the quadratic integrability of the acquired solution is achieved here . to verify this hypothesis , we have to do the integration of the @xmath0 close to @xmath6 in all of the outlined cases .
this task is still in progress .
+ for zero energy close to @xmath6 for different numbers of defects in the conical tip : influence of the @xmath1 only ( left ) and the simultaneous influence of the @xmath1 and the coulomb interaction ( right).,width=642 ] from the tip.,width=642 ] from the tip.,width=642 ] in figs .
[ compare_sc1 ] and [ compare_sc2 ] , we see the comparison of all the investigated effects which can influence the electronic structure of the graphitic nanocone . from the graphs
follows that at higher distances , the biggest significance has the effect of the @xmath1 and its influence strongly depends on the number of the defects , while the effect of the coulomb interaction ( with or without the @xmath1 ) nearly does not depend on the number of the defects . on the other hand , at the lower distance , just after the fast decrease of the @xmath0 close to @xmath6 , the effect of the @xmath1 decreases significantly and this decrease is stronger for the higher number of the defects .
the effect of the coulomb interaction in the presence of the @xmath1 decreases as well for the higher number of the defects and it approaches zero for @xmath10 . if the coulomb interaction occurs without the @xmath1 , its influence , thanks to the decrease of other effects , is getting more significant , but significant changes of its magnitude are appreciable only close to zero . +
the results described in the previous sections were acquired by adding different influences to the basic hamiltonian from @xcite , but the presented solutions are distorted by mistakes coming from the used methods of finding the numerical solution .
more precise result of the @xmath0 can be got , if we try to calculate it from the expression @xmath47 where @xmath48 is the green function corresponding to the hamiltonian ( [ hamsoc ] ) .
we can get it with the help of the perturbation calculation from the green function @xmath49 which corresponds to the hamiltonian without the influence of the @xmath1 @xcite , where the analytical solution is known .
so we can divide the hamiltonian ( [ hamsoc ] ) as @xmath50 with @xmath51 and @xmath52 here we denote @xmath53 in the perturbation scheme holds @xmath54 @xmath55 @xmath56 here , @xmath57 is specified in @xcite
. there could be a mismatch thanks to different size of @xmath49 and @xmath58 : the matrix @xmath49 as defined in @xcite is of the kind @xmath59 while the matrix @xmath58 is of the kind @xmath60 .
this discrepancy is solved in such a way that @xmath49 in the calculations is considered as @xmath61 .
we will calculate the approximation of the green function to the first and to the second order .
the first order approximation has the form @xmath62 after the substitution of all the appropriate expressions we get for given @xmath63 @xmath64}e^{{\rm i}(n - n')\varphi}\cdot\mathbb{m}_n(\varphi),\ ] ] where the matrix @xmath65 comes from the matrix multiplication @xmath66 it has the form @xmath67 @xmath68 where the coefficients @xmath69 are specified in @xcite ( eqs .
( a.4 ) , ( a.5 ) )
. we will be interested in the @xmath0 which is connected with a concrete atomic site characterized by the concrete coordinates . in this case
, we can suppose @xmath70 and because of the radial symmetry , we can suppose @xmath71 .
then , the first of the exponentials in ( [ g1sub ] ) can be removed . on the whole ,
the angular part of ( [ g1sub ] ) gives @xmath72\sin\varphi{\rm d}\varphi\,\,\,\,{\rm or}\,\,\,\ , \int\limits_0^{2\pi}\exp\left[{\rm i}(n - n')\varphi\right]\cos\varphi{\rm d}\varphi,\,\,\,\,{\rm resp.}\ ] ] or ( in the case of some parts of @xmath73 and @xmath74 ) @xmath72{\rm d}\varphi.\ ] ] the first 2 integrals are nonzero for @xmath75 only , the third is nonzero for @xmath76 .
then , for given @xmath63 , we get after the substitution into ( [ g1sub ] ) and adjusting @xmath77 @xmath78 for the purpose of the calculation of the @xmath0 , we are interested in the trace of the corresponding matrix .
since this trace has the zero value , the first order approximation to the @xmath0 is zero .
it means that the first order approximation is not sufficient to calculate the corrections to the @xmath0 following from the influence of the @xmath1 .
the particular elements of the corresponding matrix are nonzero , but it is irrelevant for us for this moment .
so we will try to find the result from the approximation to the second order . in the general case
, we can derive for the @xmath79th order of the green function expansion the recursion equation @xmath80{\rm d}r{\rm d}\varphi+\ ] ] @xmath81{\rm d}r{\rm d}\varphi.\ ] ] the analogical recursion equation holds for other indices of the green function matrix - in the general case , if we denote @xmath82 we can write schematically @xmath83 for the second order , the matrix elements @xmath84 are written in the appendix . they are expressed with the help of the triples of the integer indices @xmath85 .
we get the acceptable combinations of them from the investigation of the condition for the angular parts of @xmath86 to be nonzero : @xmath87 for @xmath88 , @xmath89 for @xmath90 .
then , we finally get for the trace of the matrix corresponding to the green function ( which is needed for the definition of the @xmath0 in ( [ ldgreen ] ) ) @xmath91,\ ] ] where @xmath92 @xmath93 @xmath94 @xmath95 now , we will make the substitution with the help of @xcite and we will integrate over @xmath96 with the help of @xcite .
however , thanks to the appearance of the multiples of 3 coefficients @xmath97 , with the respect to their definition in @xcite , the next integration over 3 variables @xmath98 is needed .
it will be performed numerically and the energy variable will be considered complex in the result .
+ we compare the results acquired by using the green function approach and with the help of the numerical method used in @xcite , where the solution for the mode @xmath4 is presented : from ( [ ldgreen ] ) follows @xmath99,\ ] ] where the lower index @xmath100 in the @xmath101 means the mentioned precision .
to calculate @xmath49 , we exploit the expression ( a.6 ) in @xcite for the calculation of @xmath102 : @xmath103,\ ] ] where @xmath104 . in the sum
, we use only the term @xmath105 for @xmath4 .
tr@xmath106 is zero for arbitrary @xmath33 , as follows from our calculations for the first order approximation .
we calculate tr@xmath107 using ( [ eqg2 ] ) . calculated using the numerical method in @xcite ( left - see also the right part of fig . 3 in @xcite ) , zeroth order approximation calculated using the relations in @xcite ( middle ) and second order approximation calculated using ( [ eqg2 ] ) ( right ) , performed for 3 defects and the values are @xmath4 , @xmath108.,title="fig:",width=207 ] calculated using the numerical method in @xcite ( left - see also the right part of fig . 3 in @xcite ) ,
zeroth order approximation calculated using the relations in @xcite ( middle ) and second order approximation calculated using ( [ eqg2 ] ) ( right ) , performed for 3 defects and the values are @xmath4 , @xmath108.,title="fig:",width=207 ] calculated using the numerical method in @xcite ( left - see also the right part of fig . 3 in @xcite ) , zeroth order approximation calculated using the relations in @xcite ( middle ) and second order approximation calculated using ( [ eqg2 ] ) ( right ) , performed for 3 defects and the values are @xmath4 , @xmath108.,title="fig:",width=207 ] if the second order approximation is sufficient , then the graph in the left part of fig .
[ ldos_green_fg ] should be given by the sum of the graphs in the middle and in the right part .
but at first glance , the graphs in the middle and in the right part have different scales , so their sum can not approach the graph in the left part .
the reason is that no normalization was considered during the calculation of the green function approach .
so , using ( [ ldos_2 ] ) , we try to evaluate the expression @xmath109,\ ] ] where @xmath110 is the normalization constant .
) evaluated using 2 different values of @xmath110 : @xmath111 ( left ) and @xmath112 ( right).,title="fig:",width=302 ] ) evaluated using 2 different values of @xmath110 : @xmath111 ( left ) and @xmath112 ( right).,title="fig:",width=302 ] in fig .
[ ldos_approx_fg ] , we see that ( [ approx ] ) will coincide with the left part of fig .
[ ldos_green_fg ] much better for the case @xmath112 .
but still it is not complete coincidence , especially for the energy values close to zero .
so it means that the next order of the green approach is needed . in the case of the finite size of the conical structure , the energy levels would be quantized and the green function approach would be modified for the case of the discrete spectrum .
the main difference in the comparison with the previous calculations is connected with the expressions ( a.4 ) and ( a.5 ) in @xcite which would be changed into the form @xmath113 @xmath114 @xmath115 where @xmath116 .
+ the discrete energy levels would be calculated on the base of an additional requirement on the wave function which should have the zero value on the border .
we use the simplest case of the nanocone without the @xmath1 presented in @xcite .
this case provides the solution of the form ( see ( 49 ) in @xcite ) @xmath117 particular parameters have this form : @xmath118 where @xmath119 is the energy of the c - c bond ( about 3 ev ) .
so we require that one of the components of the wave function has the zero value on the border : @xmath120 where @xmath121 and @xmath122 are measured in the units of @xmath119 and the length of c - c bond , respectively .
to demonstrate the character of the resulting energy levels , we will investigate the cases of @xmath123 and @xmath124 defects , i.e. @xmath125 and @xmath126 and make the choice of the parameters @xmath127 .
next , we take the value @xmath128 as the border value of @xmath16 .
then , for the particular cases , the low - energy spectrum contains these values : - : : @xmath123 defect : @xmath129 and @xmath130 or + @xmath131 and @xmath132 - : : @xmath124 defects : @xmath133 and @xmath134 or + @xmath135 and @xmath136 we see that the dependence of the energy levels on the number of the defects in the tip is not very strong . to calculate the resulting @xmath0 , the green function approach for the discrete spectrum
would be performed with the help of the summation over other modes @xmath33 .
the boundary effects influencing the electronic structure of the graphitic nanocone were investigated and we used mostly numerical methods for the corresponding calculations .
we wanted to find a quadratically integrable solution which would preserve the main features of the solution found in @xcite , especially the isolated peak for @xmath6 and @xmath10 in the case of 3 defects which could have a potential use for the construction of the probing tip in the atomic force microscopy .
we achieved this goal only partially - the found solution for the case of the simultaneous influence of the coulomb interaction and the @xmath1 in fig .
[ graph - coulomb ] ( as well as the analytical solution for the case of the coulomb interaction only from @xcite sketched in fig .
[ graph - coulomb - bezs ] ) seems to be quadratically integrable , but the mentioned isolated peak for the case of 3 defects disappeared .
we also tried to find the possible corrections caused by the considerable rise of the mass of the electron bounded on the molecular surface close to @xmath6 ( coming from the rapid curvature ) , but the resulting changes of the @xmath0 connected with this effect seem to be negligible ( regardless we use the numerical or the analytical solution ) .
the presented results of the green function approach for the continuous spectrum ( infinite size of the nanocone ) do nt correlate sufficiently with the results of the numerical approach in @xcite .
this is given by more reasons : first , we performed the calculations of the difficult integrals numerically with the help of the program mathematica , so the precision of the acquired results is limited .
secondly , we did the calculations to the second order only .
maybe we should use also the third order approach . in further calculations , we could also consider the electronic structure of the double graphitic nanocone .
the corresponding hamiltonian would differ by the presence of a new contribution which would be constant for all range of @xmath16 excluding a small interval close to the tip , where a strong influence of the interaction with the neighboring atomic orbitals caused by the curvature would increase the resulting effect .
+ + acknowledgements the work was supported by the science and technology assistance agency under contract no .
apvv-0171 - 10 , vega grant no .
2/0037/13 and ministry of education agency for structural funds of eu in the frame of project 26220120021 , 26220120033 and 26110230061 .
r. pincak would like to thank the th division at cern for hospitality .
using ( [ grf1 ] ) and ( [ grf ] ) , we get for @xmath137 @xmath138{\rm d}r_1{\rm d}r_2{\rm d}\varphi_1{\rm d}\varphi_2+\ ] ] @xmath139{\rm d}r_1{\rm d}r_2{\rm d}\varphi_1{\rm d}\varphi_2,\ ] ] where @xmath140 @xmath141 @xmath142.\ ] ] now , for the purpose of shortcut , the terms in brackets containing the combinations of @xmath143 are denoted by @xmath144 and in this notation , the next matrix elements will be @xmath145,\ ] ] @xmath146,\ ] ] @xmath147,\ ] ] @xmath148 @xmath149 @xmath150,\ ] ] @xmath151,\ ] ] @xmath152,\ ] ] @xmath153.\ ] ] for the purpose of the calculation of the @xmath0 , knowledge of the matrix block @xmath154 is also needed .
its structure is analogous ; we also get it with the help of ( [ genrec ] ) . | the boundary effects affecting the hamiltonian for the nanocone with curvature induced spin
orbit coupling were considered and the corresponding electronic structure was calculated .
these boundary effects include the spin orbit coupling , the electron mass acquisition and the coulomb interaction .
different numbers of the pentagonal defects in the tip were considered .
the green function approach into the second order was used for getting more precise results in the case of the spin orbit coupling . |
one of the most exciting developments in the recent history of quantum optics is represented by the so - called _ homodyne tomography _ , namely the homodyne detection of a nearly single - mode radiation field while scanning the phase of the local oscillator @xcite .
> from a tomographic data record , in fact , the density matrix elements can be recovered , thus leading to a complete characterization of the quantum state of the field .
this is true also when not fully efficient photodetectors are involved in the measurement , provided that quantum efficiency is larger than the threshold value @xmath0 . in homodyne tomography
a general matrix element is obtained as an expectation value over homodyne outcomes at different phases . in formula @xmath1 where @xmath2 is the probability density of the homodyne outcome @xmath3 at phase @xmath4 for quantum efficiency @xmath5 and the integral kernel is given by @xmath6 while the kernel in eq .
( [ bas2 ] ) is not even a tempered distribution , its matrix elements can be bounded functions depending on the value of @xmath5 .
this is the case of the number representation of the density matrix , for which the `` pattern function '' @xmath7 can be expressed as a finite linear combination of parabolic cylinder functions @xcite .
as it comes from the experimental average @xmath8 the tomographic determination @xmath9 for the matrix element @xmath10 is meaningful only when its confidence interval is specified .
this is defined , according to the central limit theorem , as the rms value rescaled by the number @xmath11 of data . as
@xmath12 is a complex number , we need to specify two errors , one for the real part and one for the imaginary part respectively . for the real part
one has @xmath13 ^ 2 \right\}^{1/2}}{n } \
, , \label{samp2}\end{aligned}\ ] ] where @xmath14 ^ 2 \:,\label{samp3}\end{aligned}\ ] ] and likewise for the imaginary part .
+ quantum tomography opened a fascinating perspective : in fact , there is the possibility of device - independent measurements of any field - operator , including the case of generalized observables that do not correspond to selfadjoint operators as , for example , the complex field amplitude and the phase .
the first application in this direction has been presented in ref .
@xcite where the number and the phase distributions of a low excited coherent state have been recovered from the original tomographic data record .
no error estimation was reported in ref .
@xcite , whereas an earlier analysis of the precision of such determinations has been reported in ref .
@xcite on the basis of numerical simulations .
the idea behind these papers is simple .
any field operator @xmath15 , in fact , is described by its matrix elements @xmath16 in the number representation .
then , upon a suitable truncation of the hilbert space dimension , at the maximum photon number @xmath17 , the expectation value of @xmath15 is given by the linear combination @xmath18 whereas the corresponding confidence interval is evaluated by error propagation calculus @xmath19 the whole procedure relies on two assumption , namely @xmath20 and @xmath21 which needs a more careful analysis . the condition in eq .
( [ cond1 ] ) is certainly fulfilled for some value of @xmath17 , whose determination , however , requires an _ a priori _ knowledge of the state under examination . on the other hand ,
it has been shown in ref .
@xcite that in a tomographic measurement involving @xmath11 experimental data the errors @xmath22 and @xmath23 saturate to the value @xmath24 for @xmath25 , whereas they diverge exponentially for @xmath26 .
therefore , the condition ( [ cond2 ] ) can not be fulfilled in a real experiment and we conclude that determinations of ref .
@xcite are not meaningful , as they are affected by diverging errors and are based on _ a priori _ knowledge of the state . for the same reason the analysis of ref .
@xcite is not correct , and the added noise has been largely overestimated .
in this paper we analyze the tomographic determination of field - quantities from a different perspective . by _ homodyning _ an observable @xmath15 we mean the average @xmath27(x;\phi ) \ , , \label{obs1}\end{aligned}\ ] ] of the state - independent kernel function @xmath28(x;\phi)$ ] @xcite , which allows for the determination of the expectation value @xmath29 without the detour into density matrix elements . for a hilbert - schmidt operator @xmath15 eq.([obs1 ] ) follows directly from a generalization of eq.([bas1 ] ) with @xmath28(x;\phi)=\hbox{\rm tr } \left\ { \hat a k(x-\hat x_{\phi } ) \right\}$ ] , whereas alternative approaches to derive explicit expressions of the kernel have been suggested @xcite , that here we briefly recall . starting from the identity involving trilinear products of hermite polynomials ( valid for @xmath30 even @xcite ) @xmath31 richter proved the following nontrivial formula for the expectation value of the normally ordered field operators @xcite @xmath32 which corresponds to the kernel @xmath33(x;\phi)=e^{i(m - n)\phi } \frac{h_{n+m}(\sqrt{2}x)}{\sqrt{2^{n+m}}{{n+m}\choose n}}\;.\end{aligned}\ ] ] for nonunit quantum efficiency
the homodyne photocurrent is rescaled by @xmath34 whereas the normally ordered expectation @xmath35 gets an extra factor @xmath36 .
therefore , one has @xmath37(x;\phi)=e^{i(m - n)\phi } \frac{h_{n+m}(\sqrt{2\eta}x)}{\sqrt{(2\eta)^{n+m}}{{n+m}\choose n}}\ ; , \label{rnm}\end{aligned}\ ] ] where the kernel @xmath38(x;\phi)$ ] is defined as in eq .
( [ obs1 ] ) , but now with the experimental probability distribution @xmath2 for nonunit quantum efficiency @xmath5 . from eq .
( [ rnm ] ) by linearity one can obtain the kernel @xmath39(x;\phi)$ ] for any operator function @xmath40 that admits a normal ordered expansion @xmath41 one obtains @xmath42(x;\phi)&=&\sum_{s=0}^{\infty } \frac{h_s(\sqrt{2\eta}x)}{s!(2\eta)^{s/2}}\sum_{nm=0}^{\infty } f^{(n)}_{nm}e^{i(m - n)\phi}n!m!\delta_{n+m , s}\nonumber \\ & = & \sum_{s=0}^{\infty}\frac{h_s(\sqrt{2\eta}x)i^s}{s ! ( 2\eta)^{s/2}}\frac{d^s}{dv^s}\bigg|_{v=0}\!\!\!\ ! { \cal f}[\hat f](v;\phi),\label{123}\end{aligned}\ ] ] where @xmath43(v;\phi)=\sum_{nm=0}^{\infty}f^{(n)}_{nm } { { n+m}\choose m}^{-1}(-iv)^{n+m}e^{i(m - n)\phi}\;.\label{ff}\end{aligned}\ ] ] continuing from eq . ( [ 123 ] ) one obtains @xmath42(x;\phi)=\exp\left(\frac{1}{2\eta}\frac{d^2}{dv^2}+
2ix\frac{d}{dv}\right)\bigg|_{v=0}{\cal f}[\hat f](v;\phi)\;,\end{aligned}\ ] ] and finally @xmath42(x;\phi)=\int_{-\infty}^{+\infty } \frac{dw}{\sqrt{2\pi\eta^{-1}}}e^{-\frac{\eta}{2}w^2}{\cal f}[\hat f ] ( w+2ix;\phi)\;.\label{generalf}\end{aligned}\ ] ] in summary , the operator @xmath40 possesses a tomographic kernel @xmath39(x;\phi)$ ] if the function @xmath44(v;\phi)$ ] in eq .
( [ ff ] ) grows slower than @xmath45 for @xmath46 .
in addition , as we can assume that @xmath2 goes to zero faster than exponentially at @xmath47 , the average in eq .
( [ obs1 ] ) is meaningful for the integral in eq .
( [ generalf ] ) growing at most exponentially for @xmath47 . in the next section
we will consider the tomographic determination of four relevant field quantities : the field intensity , the real field or quadrature , the complex field , and the phase , for all of which the above conditions are satisfied .
as already mentioned in the previous section the tomographic measurement of the quantity @xmath15 is _ defined _ as the average @xmath48 of the kernel @xmath49 ( x,\phi ) $ ] over the homodyne data . a convenient measure for the precision of the measurement
is given by the confidence interval @xmath50 which , being @xmath51 a real quantity , is given by @xmath52 , where @xmath53(x,\phi)}= \int^{\pi}_{0 } \frac{d\phi}{\pi}\int_{-\infty}^{\infty } \!\ !
dx\:p_{\eta } ( x,\phi)\ : { \cal r}^2_{\eta}[\hat a](x,\phi)\ : .
\label{ysquared}\end{aligned}\ ] ] when the quantity @xmath15 can also be directly measured by a specific setup it makes sense to compare tomographic precision @xmath54 with the corresponding fluctuations @xmath55 .
notice that , when we deal with @xmath56 the noise @xmath57 is larger that the quantum fluctuations due to smearing effect of nonunit quantum efficiency . as we will see , the tomographic measurement is always more noisy than the corresponding direct measurement for any observable , and any quantum efficiency @xmath5 . however , this is not surprising , in view of the larger amount of information retrieved in the tomographic measurement compared to the direct measurement of a single quantity .
+ in table [ t : one ] we report the tomographic quantities @xmath51 the field - observables examined . before going into details of each observable , we mention a useful formula for evaluating confidence intervals .
these are obtained by averaging quantities like @xmath58(x,\phi)= \label{er1 } e^{2i\phi(m - n ) } \frac{h^2_{n+m}\left(\sqrt{2\eta}x\right ) } { ( 2\eta)^{(n+m ) } \left(\begin{array}{c}m+n \\ m\end{array}\right)^2 } \:.\end{aligned}\ ] ] by means of the following identity for the hermite polynomials @xcite @xmath59 we arrive at @xmath60(x,\phi ) = e^{2i\phi(m - n ) } \frac{n!^2 m!^2}{\eta^{m+n } } \sum_{k=0}^{m+n}\frac{(2k)!\eta^k}{k!^4 ( n+m - k)!}{\cal r}_{\eta}[a^{\dag k } a^k](x,\phi ) \label{er3}\ : , \end{aligned}\ ] ] which expresses the generic square kernel @xmath61(x,\phi)$ ] in terms of `` diagonal '' kernels @xmath62(x,\phi)$ ] only .
photodetection is the direct measurement of the field - intensity . for a single - mode of the radiation field it corresponds to the number operator @xmath63 . for nonunit quantum efficiency @xmath5 at the photodetectors ,
only a fraction of the incoming photons is revealed , and the probability of detecting @xmath64 photons is given by the bernoulli convolution @xmath65 @xmath66 being the actual photon number distribution of the mode under examination .
one considers the reduced photocurrent @xmath67 which is the quantity that traces the photon number , namely it has the same mean value @xmath68 where we introduced the shorthand notation @xmath69 . on the other hand
, one has @xmath70 in eq .
( [ pho4 ] ) @xmath71 denotes the intrinsic photon number variance .
the term @xmath72 represents the noise introduced by inefficient detection . the tomographic kernel that traces the photon number
is given by the phase - independent function @xmath73 .
with the help of eq .
( [ er3 ] ) we can easily evaluate its variance , namely @xmath74 the difference between @xmath75 and @xmath76 defines the noise @xmath77 $ ] added by tomographic method in the determination of the field intensity @xmath78 = \frac{1}{2}\left [ \langle\widehat{n^2}\rangle+ \bar{n}\left(\frac{2}{\eta}-1 \right ) + \frac{1}{\eta^2}\right ] \label{pho6}\:.\end{aligned}\ ] ] the noise @xmath77 $ ] added by the tomographic measurement is always a positive quantity and largely depends on state under examinations . for coherent states we consider the noise - ratio @xmath79 which is minimum for @xmath80 . for a single mode light - beam the electric field is proportional to a field quadrature @xmath81 , which is just traced by homodyne detection at fixed zero - phase with respect to the local oscillator
the tomographic kernel , that traces the mean value @xmath82 is given by @xmath83(x,\phi ) = 2 x \cos\phi$ ] , independently on @xmath5 , whereas the square kernel @xmath84(x,\phi ) = 4 x^2 \cos^2\phi$ ] can be rewritten as @xmath85(x,\phi ) + { \cal r}[a^{\dag 2}](x,\phi)\right ] + { \cal r}[a^{\dag } a](x,\phi ) + \frac{1}{2\eta } \label{fie1}\;.\end{aligned}\ ] ] the confidence interval is thus given by @xmath86 + \bar{n } + \frac{1}{2\eta } - \left\langle \frac{a+a^{\dag}}{2}\right\rangle^2 = \langle\widehat{\delta x^2 } \rangle + \frac{1}{2}\bar{n } + \frac{2-\eta}{4\eta } \label{fie11}\ ; , \end{aligned}\ ] ] @xmath87 being the intrinsic quadrature fluctuations . for coherent states
( [ fie11 ] ) reduces to @xmath88 \label{fie12}\ ; , \end{aligned}\ ] ] the tomographic noise in eq .
( [ fie11 ] ) has to be compared with the rms variance of homodyne detection for nonunit quantum efficiency .
this is given by @xmath89 for coherent states eq .
( [ fie20 ] ) becomes @xmath90 the added noise results @xmath91= \frac{1}{2}\left[\bar{n}+\frac{1}{2\eta}\right ] \label{fie2}\ ; , \end{aligned}\ ] ] whereas the noise - ratio for coherent states is given by @xmath92 and increases with the scaled intensity @xmath93 .
the detection of the complex field amplitude of a single - mode light - beam is represented by the generalized measurement of the annihilation operator @xmath94 .
the tomographic kernel for @xmath94 is given by the complex function @xmath95(x,\phi ) = 2x \exp(i\phi)$ ] .
to evaluate the precision of the measurement one has to consider the noise of a complex random variable .
generally there are two noises @xmath96 corresponding to the eigenvalues of the covariance matrix .
( [ er3 ] ) one has @xmath97(x,\phi )
= e^{i2\phi } \left[\frac{1}{\eta}+2{\cal r}_{\eta}[a^{\dag}a ] ( x,\phi ) \right ] = \frac{e^{i2\phi}}{\eta } + { \cal r}_{\eta}[a^2 ] ( x,\phi ) \label{amp2}\:.\end{aligned}\ ] ] and @xmath98(x,\phi)\right|^2 = \frac{1}{\eta } \left[1 + 2\eta { \cal r}_{\eta}[a^{\dag}a ] ( x,\phi)\right ] \label{amp3}\ : , \end{aligned}\ ] ] which lead to @xmath99 \label{amp4}\:,\end{aligned}\ ] ] because @xmath100 for all states .
the optimal measurement of the complex field @xmath94 , corresponding to the joint measurement of any pair of conjugated quadratues @xmath101 and @xmath102 can be accomplished in a number of different ways : by heterodyne detection @xcite , eight - port homodyne detection @xcite , or by six - port homodyne detection @xcite . in such devices
each experimental event @xmath103 in the complex plane consists of a simultaneous detection of the two commuting photocurrents @xmath104 and @xmath105 , which in turn trace the pair of field - quadratures .
the probability distribution is represented by the generalized wigner function @xmath106 with ordering parameter @xmath107 related to the quantum efficiency as @xmath108 .
the precision of such measurement is defined likewise eq .
( [ amp1 ] ) as follows @xmath109 \label{amp41}\:,\end{aligned}\ ] ] where @xmath110 > from eqs .
( [ amp41 ] ) and ( [ amp5 ] ) we have @xmath111 \label{amp51}\:,\end{aligned}\ ] ] the noise added by quantum tomography is thus simply given by @xmath112 = \frac{1}{2}\bar{n } \label{amp6}\ : , \end{aligned}\ ] ] which is independent on quantum efficiency .
+ for a coherent state we have @xmath113 \quad \hbox{\rm and } \quad \langle\widehat{\delta a^2}\rangle _ { \eta}= \frac{1}{2\eta } \label{amp7}\ : , \end{aligned}\ ] ] and the noise ratio is given by @xmath114 the canonical description of the quantum optical phase is given by the probability operator measure @xcite @xmath115 which defines a phase operator @xcite through the relation @xmath116 in principle , a comparison between homodyne tomography and direct determination of the phase would require from one side the average of the kernel corresponding to the operator @xmath117 , and from the other side the direct experimental sample of the operator @xmath117 .
however , such a comparison would be purely academic , as there is no feasible setup achieving the optimal measurement ( [ pha1 ] ) . for this reason ,
here we consider the heterodyne measurement of the phase , and compare it with the phase of the tomographic kernel for the corresponding field operator @xmath94 , i.e. @xmath118 .
notice that the phase @xmath51 is not just the given local oscillator phase , because x has varying sign .
hence averaging @xmath51 is not just the trivial average over the scanning phase @xmath4 .
the probability distribution of such kernel variable can be easily obtained by the following identity @xmath119 which implies @xmath120 the precision in the tomographic phase measurement is given by the rms variance @xmath75 of the probability ( [ pha5 ] ) . in the case of a coherent state @xmath121\right\ } \label{pha6}\ : , \end{aligned}\ ] ] which approaches a `` boxed '' distribution in @xmath122 $ ] for large intensity .
we compare the tomographic phase measurement with its heterodyne detection , namely the phase of the detected complex field @xmath94 .
the outcome probability distribution is the marginal distribution of the generalized wigner function @xmath106 ( @xmath108 ) integrated over the radius @xmath123 whereas the precision in the phase measurement is given by its rms variance @xmath124 .
we are not able to give a closed formula for the added noise @xmath125=\overline{\delta w_{\eta}}^2-\overline{\delta\phi}_{\eta}^2 $ ] . however , for high excited coherent states @xmath126 and @xmath127 .
the asymptotic noise - ratio is thus given by @xmath128 a comparison for low excited coherent states can be performed numerically . the noise ratio @xmath129 ( expressed in db ) is shown in fig .
[ f : pha ] for some values of the quantum efficiency @xmath5 .
it is apparent that the tomographic determination of the phase is more noisy than the heterodyne one also in this low - intensity regime .
homodyne tomography provides a complete characterization of the state of the field . by averaging suitable kernel functions it is possible to recover the mean value of essentially any desidered field - operator . in this paper
we analyzed the determination of the most relevant observables : intensity , real and complex field , phase .
we have shown that these determinations are affected by noise , which is always larger than the corresponding one from the direct detection of the considered observables . in table
[ t : two ] a synthesis of our results is reported .
+ we have considered the ratio between the tomographic and the direct - measurement noises .
this is an increasing function of mean photon number @xmath130 , however scaled by the quantum efficiency @xmath5 .
therefore homodyne tomography turns out to be a very robust detection scheme for low quantum efficiency . in fig .
[ f : cmp ] the noises ratio ( in db ) for all the considered quantities are plotted for unit quantum efficiency versus @xmath130 : this plot has to compared with fig .
@xcite , where the tomographic errors were largely overestimated .
+ in conclusion , homodyne tomography adds larger noise for highly excited states , however , it is not too noisy in the quantum regime of low @xmath130 .
it is then a matter of convenience to choose between a direct measurement and homodyne tomography , as the former is the most precise measurement of the desidered quantity , whereas the latter represents the best compromise between the conflicting requirements of a precise and complete measurement of the state of radiation .
99 d. t. smithey , m. beck , m. g. raymer , a. faridani , phys .
* 70 * , 1244 ( 1993 ) .
g. m. dariano , c. macchiavello , m. g. a. paris , phys .
rev . a*50 * 4298 ( 1994 ) .
g. m. dariano , u. leonhardt , h. paul , phys .
rev . a*52 * , r1801,(1995 ) .
g. m. dariano , in in _ concepts and advances in quantum optics and spectroscopy of solids _ , ed . by t. hakioglu and a. s. shumovsky .
( kluwer , amsterdam 1996 d. t. smithey , m. beck , j. cooper , and m. g. raymer , phys . rev . a * 48 * , 3159 ( 1993 ) . g. m. dariano , c. macchiavello , m. g. a. paris , phys . lett . a*195 * , 31 ( 1994 ) . g. m. dariano , c. macchiavello , n. sterpi , j. mod . opt .
, to appear g. m. dariano , tokyo i. s. gradshteyn , i. m. ryzhik , _ table of integral , series , and product _ , ( academic press , 1980 ) .
richter , phys .
a * 221 * 327 ( 1996 ) .
a. orlowsky , a. wnsche , phys .
rev . a*48 * 4617 ( 1993 ) .
j. h. shapiro , s. s. wagner , ieee j. quantum electron .
qe*20 * , 803 ( 1984 ) ; h. p. yuen , j. h. shapiro , ieee trans .
theory it*26 * , 78 ( 1980 ) .
walker , j.e .
carrol , opt .
quantum electr . *
18 * , 355(1986 ) ; n. g. walker , j. mod
. opt . * 34 * , 15 ( 1987 ) .
y. lay , h. a. haus , quantum opt . * 1 * , 99 ( 1989 ) .
g. m. dariano , m. g. a. paris , phys . rev . * 49 * 3022 ( 1994 ) .
a. zucchetti , w. vogel , d .-
welsch , phys .
rev . a*54 * 856 ( 1996 ) m. g. a. paris , a. chizhov , o. steuernagel , opt . comm . * 134 * , 117 ( 1997 ) .
a. s. holevo , _ probabilistic and statistical aspects of quantum theory _ ( north - holland publishing , amsterdam , 1982 ) . c. w. helstrom , _ quantum detection and estimation theory _
( academic press , new york , 1976 ) .
popov v p and yarunin v s 1992 _ j. mod .
opt . _ * 39 * 1525 | homodyne tomography provides a way for measuring generic field - operators . here
we analyze the determination of the most relevant quantities : intensity , field , amplitude and phase .
we show that tomographic measurements are affected by additional noise in comparison with the direct detection of each observable by itself .
the case of of coherent states has been analyzed in details and earlier estimations of tomographic precision are critically discussed . |
in this paper , we obtain a reformulation of the four color theorem in terms of signed permutations ( theorem [ fc ] ) .
signed permutations are standard words ( that is with no letter repeated ) @xmath4 of length @xmath5 on the alphabet @xmath6 which do not contain any pair @xmath7 the barred ( resp .
unbarred ) letters are interpreted as negative ( resp .
positive ) integers and we write @xmath8 for the word on @xmath9 obtained by erasing the bars which appear on the letters of @xmath10we first describe a combinatorial procedure which attaches to each permutation @xmath11 ( considered as a standard word on a given totally ordered alphabet @xmath12 ) a triangulation @xmath13 of a fixed @xmath2-gon @xmath14
. the fibers of this association coincide with the sylvester classes defined in @xcite .
our construction is close to that used by reading in @xcite and the sylvester congruence on _ standard _ words is in fact a special case of cambrian congruence .
nevertheless , we have chosen to give an independent exposition of the occurrence of the sylvester relations in the context of triangulations .
this makes the paper self - contained and permits us to expose precisely the results which are required in our reformulation of the four color theorem in terms of signed permutations .
this connection with the results of @xcite motivates an additional reformulation of the four color theorem in terms of the geometry of the associahedron .
we also introduce colored triangulations obtained by associating a color ( considered as a letter of a totally ordered alphabet @xmath15 with @xmath16 colors ) to each face ( or to each vertex ) of a given triangulation of @xmath14 .
the map @xmath17 admits a natural extension @xmath18 defined from the set of words @xmath4 with letters in @xmath15 to a particular subset of colored triangulations we have called simple .
we show that the fiber @xmath19 coincides with the sylvester class of @xmath4 ( theorem [ th_sylvt ] ) .
this implies that the simple colored triangulations can be regarded as combinatorial objects analogous to binary search trees .
the map @xmath18 can be interpreted as an insertion scheme for simple triangulations .
this gives an alternative to knuth s insertion algorithm on binary search trees .
_ signed triangulations _ are colored triangulations with @xmath20 .
to each triangulation @xmath21 of @xmath14 corresponds the sylvester class @xmath22this permits us to associate to each signed triangulation @xmath23 defined as a signing @xmath24 of @xmath21 , the set @xmath25_{s}$ ] of signed permutation words @xmath4 such that @xmath26 . to obtain our reformulation of the four color theorem , we establish that for each signed flip ( defined as a particular diagonal flip ) in @xmath27 one can find two words @xmath28 and @xmath29 in @xmath25_{s}$ ] such that@xmath30{rcl}% w_{1 } & = & u\ \alpha\ \gamma\ v\\ w_{2 } & = & u\ \overline{\gamma}\ \overline{\alpha}\ v \end{array}\ ] ] where @xmath31 are letters with the same sign , @xmath32 are factors of @xmath28 and @xmath33 does not contain any letter @xmath34 such that @xmath35 we also study the combinatorial problem of describing the graph obtained from a colored triangulation @xmath27 by computing successive flips .
when no condition is imposed on the flip , it is well known that the flip graph of the ( ordinary ) triangulations @xmath21 is connected and contains all the triangulations of @xmath36 in addition to signed flips , we also consider in this paper switched flips .
they are flips which preserve the coloring and for which a flip operation is authorized when the two faces considered have distinct colors . for any @xmath37 such that @xmath38 , let @xmath39 be the frobenius subgroup of @xmath1 defined by @xmath40we prove that the map @xmath18 is a morphism from the cayley graph of @xmath41 to the graph whose vertices are the simple colored triangulations associated to @xmath42 connected by an edge when they differ by a switched flip .
so the graph defined from a simple colored triangulation by applying switched flips contains all the simple colored triangulations which have the same coloring and thus , is connected .
the paper is organized as follows .
section @xmath43 is devoted to the combinatorial background on triangulations , colored triangulations and flip operations we need in the sequel . in section @xmath44 we introduce the maps @xmath17 and @xmath18 and prove that their fibers coincide with sylvester classes .
the reformulations of the four color theorem using respectively the signed permutations and the geometry of the associahedron are given in section @xmath45 .
finally , we study in section @xmath46 the flip graphs generated by colored triangulations when restrictive conditions are imposed on the flips .
consider an integer @xmath47 and @xmath48 a subset of @xmath49 of size @xmath50we denote by @xmath51 the augmented set @xmath52 , totally ordered as follows : @xmath53 each integer @xmath54 has a predecessor and a successor in @xmath51 that we denote @xmath55 and @xmath56 ( or @xmath57 and @xmath58 , for short ) , respectively .
thus , @xmath59 is a 3-element interval in @xmath51 .
let @xmath60 be a convex @xmath2-gon , with vertices labelled by the elements of @xmath51 in a clockwise increasing way . by a _
@xmath21 of @xmath14 , we mean a plane graph with the @xmath61 vertices and @xmath61 edges of @xmath14 , and with @xmath62 additional edges , called _ diagonals _ , which subdivide the inner face of @xmath14 into @xmath5 triangles , called the _ faces _ of @xmath21 .
we denote by @xmath63 the set of triangulations of the polygon @xmath64it is well known that the cardinality of @xmath63 is equal to the catalan number @xmath65 by a slight abuse of notation , we shall make no distinction between a vertex @xmath33 of @xmath14 and its label in @xmath51 . in other words
, we consider the polygon @xmath14 and its triangulations @xmath21 as graphs with vertex set @xmath66 recall that the _ degree _ of a vertex @xmath33 in a graph @xmath67 is the number of edges of @xmath67 which are incident to @xmath68 now , in any triangulation @xmath21 of the polygon @xmath14 , as above , the degree of each vertex @xmath33 is at least 2 , because of the two edges of @xmath14 which are incident to @xmath33 .
an _ ear _ in @xmath21 is a vertex of degree exactly @xmath43 , i.e. a vertex incident to zero diagonals of @xmath21 .
thus , an ear @xmath69 belongs to exactly one face of @xmath21 , whose three vertices are @xmath70 ( more generally , a vertex of degree @xmath71 in @xmath21 belongs to exactly @xmath72 faces of @xmath21 . ) it is easy to see that every triangulation of @xmath14 contains at least two ears , and that no two ears are adjacent if @xmath73 . by convention , we shall label each face @xmath74 of a triangulation @xmath21 by its middle vertex .
in other words , if @xmath75 are the three vertices of the face @xmath74 , we shall label @xmath74 by @xmath76 .
this labelling gives a bijection between the faces of @xmath21 and the set @xmath77 ; there are no faces labelled @xmath78 of @xmath79 .
figure2.eps + figure 1 : labelling of the faces [ figure2 ] consider now a totally ordered set @xmath80 , which will be referred to in the sequel as the _ set of colors_. given a triangulation @xmath21 and @xmath81 a @xmath5-tuple of colors , we call _ colored triangulation _ the triangulation @xmath27 obtained by replacing in @xmath21 each label @xmath82 by the pair @xmath83 the faces of @xmath27 are colored by associating the color @xmath84 to the face labelled @xmath85 by the previous procedure .
the vertex @xmath85 and the face @xmath86 in the triangulation @xmath27 are hence colored by the color @xmath87 .
note that the additional vertices labelled by @xmath78 and @xmath79 are not colored .
we write simply @xmath21 for the underlying triangulation associated to @xmath27 .
denote by @xmath88 the set of triangulations of @xmath14 colored by @xmath89when @xmath90 we set @xmath91 ( with @xmath92 ) and call _ signed triangulations _
the colored triangulations on @xmath93 we will say that @xmath24 is an _ increasing coloring _ when @xmath94 [ de_coltrian]the colored triangulation @xmath27 is called simple if it verifies the following conditions : * @xmath24 is an increasing coloring ; * there are no inner diagonals in @xmath27 connecting two vertices with the same color ; * when two consecutive vertices @xmath85 and @xmath95 are colored with the same color , the third vertex @xmath96 of the face they define verifies @xmath97 * remarks . * @xmath98 in a simple colored triangulation , there is no face with three vertices of the same color .
similarly , if a face has two vertices colored with the same color , these vertices must be consecutive .
@xmath99 in a colored triangulation , the coloring of the vertices determines that of the faces .
conversely , the coloring of the faces gives the coloring of the vertices since each face of a triangulation has the color of its middle vertex .
so the colored triangulation @xmath27 is characterized by @xmath21 and one of these two colorings . @xmath100%
{ figure3b.eps}% } % % endexpansion% % tcimacro{\frame{itbpf}{5.0918cm}{4.8084cm}{0cm}{}{}{figure3a.eps}% % { \special { language " scientific word " ; type " graphic " ; % maintain - aspect - ratio true ; display " usedef " ; valid_file " f " ; % width 5.0918 cm ; height 4.8084 cm ; depth 0 cm ; original - width 3.9695 in ; % original - height 3.7464 in ; cropleft " 0 " ; croptop " 1 " ; cropright " 1 " ; % cropbottom " 0 " ; filename ' figure3a.eps';file-properties " xnpeu";}}}% % beginexpansion \raisebox{-0cm}{\includegraphics [ height=4.8084 cm , width=5.0918 cm ] % { figure3a.eps}% } % % endexpansion \label{vc}\\ \text{figure 2 : non simple and simple colored triangulations for } % \mathcal{c}=\{a < b < c\}\nonumber\end{gathered}\ ] ] the set of simple triangulations colored by @xmath24 will be denoted by @xmath101 consider @xmath102 a triangulation of the polygon @xmath103 the triangulation @xmath21 may be transformed into another one by a diagonal flip .
the _ diagonal flip _ , or _
flip _ for short , of a diagonal @xmath71 in @xmath21 is the following operation : in the quadrilateral @xmath104 formed by the two faces of @xmath21 adjacent to @xmath105 , remove @xmath71 and replace it by the opposite diagonal @xmath106 the result is a new triangulation @xmath107 of @xmath14 ( see figure below ) .
it is well known that given any triangulations @xmath108 and @xmath109 of @xmath110 there exists a sequence of diagonal flips transforming @xmath108 into @xmath111the graph on @xmath63 obtained by joining two triangulations which differ by exactly one flip is called the _ flip graph _
( see @xcite ) .
we will denote it by @xmath112there are many other labellings of the flip graph by objects enumerated by the catalan numbers , such as binary trees , parenthesizations , etc .
we now introduce various types of _ colored flips _ on colored triangulations of @xmath14 , by imposing special constraints on the two faces adjacent to the diagonal being flipped .
the notion of signed flips plays a key role in our reformulation of the four color theorem .
* a _ signed flip _ in a signed triangulation is a diagonal flip such that the two faces adjacent to the flipped diagonal have equal signs , which are changed after the flip . *
a _ homogeneous flip _ in a colored triangulation is a diagonal flip such that the two faces adjacent to the flipped diagonal have the same color , which are preserved after the flip .
* a _ switched flip _ is a diagonal flip which preserves the vertex colors and such that the two faces adjacent to the flipped diagonal have different colors .
such flips will be referred to in the sequel as _ restricted flips_.
figure6.eps[fig6 ] figure6b.eps + figure 3 [ fig3 ]
as above , let @xmath113 be a subset of @xmath114 . as usual , we denote by @xmath1 the symmetric group of rank @xmath5 . by realizing @xmath1 as the permutation group of
@xmath115 one can identify each @xmath11 with a standard word @xmath116 on @xmath117 we shall now define a map @xmath118 which , being surjective , will allow us to represent each triangulation of @xmath14 by a suitable word in @xmath1 . consider @xmath116 a permutation of @xmath119the associated triangulation @xmath13 of @xmath14 is defined by adding @xmath62 noncrossing diagonals to @xmath14 with the following algorithm : 1 . in @xmath120
, join the two neighbors of @xmath121 , i.e. @xmath122 and @xmath123 , by a diagonal .
2 . for each integer @xmath124 consider the polygon @xmath125 obtained from @xmath126 by deleting the vertex @xmath127 and the edges connected to it , then joining the two neighbors of @xmath128 in the polygon @xmath125 by a diagonal . + figure4.eps + figure 4 : the triangulation @xmath129[fig4 ] one easily verifies that the plane graph obtained when the procedure terminates is a triangulation @xmath21 of @xmath36 * remark .
* in @xcite , the author uses a similar combinatorial map , denoted @xmath130 , from the symmetric group to the set of triangulations .
the definitions of the maps @xmath17 and @xmath130 are quite different , but one can verify that for any @xmath131 one has @xmath132 we shall now prove that each triangulation @xmath102 can be represented by a suitable word @xmath11 .
[ cutt]the map @xmath133 is surjective .
we proceed by induction on @xmath50 the lemma is trivial for @xmath134 .
assume @xmath73 and the statement true for @xmath62 .
let @xmath135 and let @xmath21 be a triangulation of the @xmath2-gon @xmath14 with vertex set @xmath136 since @xmath21 contains at least two non - adjacent ears , some vertex @xmath137 must be an ear . denote by @xmath107 the triangulation obtained by _ cutting the ear @xmath85 _ in @xmath21 , i.e. by deleting in @xmath21 the unique face containing the ear @xmath85 . then @xmath107 is a triangulation of a convex @xmath138-gon @xmath139 on the vertex set @xmath140 , so that @xmath141 . by the induction hypothesis , @xmath142 where @xmath143 is a permutation of the set @xmath144 .
denoting @xmath145 , we have @xmath146 by construction .
we shall refer to the procedure used in the above proof as the _ cutting ear procedure_. the words in @xmath1 obtained with this procedure will be called the _ readings _ of the triangulation @xmath147from the previous lemma we have @xmath146 for any reading @xmath0 of @xmath147the _ canonical reading _ of @xmath21 is the reading obtained by considering the cutting ear procedure for which the vertex with the greatest label is deleted at each step .
this is equivalent to say that the readings of @xmath21 are the words in @xmath148 and the canonical reading of @xmath21 is the greatest reading for the lexicographic order ( see example [ read ] ) .
we shall now slightly generalize the preceding construction , by associating a suitably colored triangulation to a word where letters are allowed to be repeated .
consider a word @xmath4 of length @xmath5 on the alphabet @xmath149 .
the _ evaluation _ of @xmath4 is the @xmath16-uple @xmath150 where for any @xmath151 @xmath152 is the number of occurrences of the color @xmath153 in @xmath10we write for short @xmath154 we denote by @xmath155 the set of words of length @xmath5 and evaluation @xmath42 on the alphabet @xmath15 .
the _ standardization _ of @xmath4 on @xmath156 will be denoted by @xmath157 .
recall that @xmath157 is obtained by labelling from @xmath158 to @xmath159 the occurrences of the color @xmath160 reading from left to right , then from @xmath161 to @xmath162 the occurrences of @xmath163 and so on .
( see the example below . ) for any @xmath11 considered as a word on @xmath164 of length @xmath165 one associates the sequence @xmath166 where @xmath167 is the longest increasing sequence in @xmath0 starting at @xmath158 of successive letters read from left to right , @xmath168 the longest increasing sequence in @xmath0 starting at @xmath169 of successive letters and so on .
set @xmath170 , where @xmath171 is the length of @xmath172 for any @xmath173 and set @xmath174 .
then , for any word @xmath175 , its standardization std@xmath176 belongs to @xmath177 , and the standardization map @xmath178 is a bijection .
the corresponding inverse map is called the _ destandardization _ and denoted by @xmath179 .
suppose @xmath180 and @xmath181for @xmath182 we have@xmath183{c}% $ \mathrm{eval(}w)=(2,3,2,1),\text { } \mathrm{std}(bacbbacd)=31645278,$\\ $ \delta(31645278)=(12,345,678),$ $ l(\sigma)=(2,3,3)$. \end{tabular } \right.\ ] ] in particular @xmath184 and @xmath185 .
to each @xmath5-tuple @xmath186 we associate the coloring@xmath187 our purpose now is to define a map @xmath188 where @xmath189 is the set of simple colored triangulations of @xmath14 with coloring @xmath190 . to the word @xmath191 we associate the colored triangulation @xmath192 , where @xmath193 .
let @xmath0 be a word in @xmath1 on the alphabet @xmath194 and let @xmath195 be the associated triangulation of the polygon @xmath14 with vertex set @xmath196 .
for each @xmath197 , denote by @xmath198 the third vertex of the unique face of @xmath21 containing the edge @xmath199 . then 1 .
the letter @xmath85 is on the left of the letter @xmath95 in @xmath0 if and only if @xmath97 2 . if @xmath0 contains the increasing sequence @xmath200 with @xmath201 in its left to right reading , then @xmath202 3 . for any word @xmath4 on @xmath15
, @xmath203 is a simple colored triangulation .
@xmath204 suppose that @xmath85 is on the left of the letter @xmath95 in @xmath205 .
assume @xmath206 .
then , when constructing @xmath195 step - by - step , the vertex @xmath85 is an ear in the subtriangulation @xmath207 , and its successor in @xmath208 is @xmath209 since the unique face of @xmath107 containing @xmath85 is @xmath210 , we must have @xmath211conversely , assume that @xmath85 is on the right of @xmath95 in @xmath0 .
then @xmath95 is an ear in some subtriangulation @xmath107 of @xmath21 , and its predecessor in the vertex set @xmath212 of @xmath107 must be @xmath213 .
thus @xmath96 is the successor of @xmath95 in @xmath212 , and therefore @xmath214 .
this proves assertion @xmath204 .
@xmath215 we know by @xmath204 that @xmath216 . by reasoning inductively on @xmath217
, it suffices to show that @xmath218 .
as above , the vertex @xmath85 is an ear in some subtriangulation @xmath107 of @xmath21 with vertex set @xmath219 , and @xmath96 is the predecessor of @xmath85 in @xmath212 . since @xmath220 is the predecessor of @xmath95 in some subset of @xmath212 , and
since @xmath221 ( as there is an edge @xmath222 but no edge @xmath223 ) , we must have @xmath218 as claimed .
@xmath224 note first that by definition of @xmath225 @xmath203 is an increasing colored triangulation .
set @xmath226 and @xmath195.@xmath227consider a face @xmath74 of @xmath203 with two vertices colored by @xmath228 in @xmath21 , @xmath74 has two vertices labelled by @xmath229 and @xmath230 such that @xmath231 .
all the letters @xmath232 with @xmath233 label vertices colored by @xmath234 in @xmath235thus they correspond to letters @xmath234 in @xmath4 . since @xmath226
, they must appear in increasing order in the left to right reading of @xmath236 suppose @xmath237 , that is @xmath238 is an inner diagonal of @xmath21 .
then @xmath239 appears in @xmath0 on the right of @xmath240when @xmath13 is constructed , a diagonal joining the vertex labelled @xmath239 to a vertex labelled by @xmath241 @xmath239 is drawn .
since there are no intersections between the inner diagonals of @xmath242 this gives a contradiction for @xmath74 is a face with an edge joining @xmath229 and @xmath230 and @xmath243this means that @xmath244 .
by @xmath245 we obtain that the third face of @xmath74 is labelled by @xmath246 thus @xmath203 is a simple colored triangulation .
figure5.eps + figure 5 : the simple triangulation @xmath247[fig5 ] consider @xmath248 ( see ( [ epsilon(mu)])).@xmath227the readings of the simple colored triangulation @xmath249 are the words obtained by applying cutting ear procedures on @xmath21 and by forming the words of @xmath155 obtained from the successive colors of the ears deleted instead of their labels . since the cutting ear procedure depends only of the inner diagonals of the triangulation @xmath21 and not on the labels or colorings of its faces , there is a one - to - one correspondence between the readings of @xmath249 and those of @xmath21 .
moreover by definition of @xmath225 the readings of @xmath21 coincide with the standardized of the readings of @xmath250 the canonical reading of @xmath249 is the greatest reading of @xmath249 for the lexicographic order . by proceeding as in lemma [ cutt ] , we prove that for any reading @xmath4 of @xmath251 we have @xmath252 in particular the map @xmath18 is surjective .
[ read]the readings of the triangulation obtained in figure 4 are the permutations @xmath253 .
thus the readings of the colored triangulation obtained in the above figure are the words @xmath254 .
the map @xmath18 can alternatively be thought of as an insertion scheme for the simple colored triangulations . indeed consider a simple colored triangulation @xmath27 .
suppose that the vertices of @xmath21 are indexed by the integers of the set @xmath255 .
let @xmath256 be a positive integer such that @xmath257 and @xmath258 .
consider the pair @xmath259 where @xmath260 there exists a unique vertex @xmath33 in @xmath27 which is not colored by @xmath234 ( thus is colored by @xmath261 since @xmath24 is an increasing coloring ) and such that @xmath262 is colored by @xmath263 the insertion of @xmath259 in @xmath27 ( denoted @xmath264 for short ) is defined as follows : * add a vertex @xmath265 indexed by @xmath259 between the vertices @xmath33 and @xmath266 in @xmath267 * draw the two edges joining @xmath265 to @xmath33 and @xmath265 to @xmath266 .
* color the face defined by @xmath268 and @xmath266 with @xmath263 1 .
the output of the insertion @xmath264 is a simple colored triangulation .
2 . for any word
@xmath269 on @xmath155 , we have @xmath270 where @xmath271 and @xmath272 is the triangulation with one edge joining the vertices
@xmath78 and @xmath273 @xmath204 if @xmath74 is a face of @xmath264 with two vertices colored by the same color @xmath274 @xmath74 is face of @xmath27 or @xmath275 in the first case , the third vertex of @xmath74 is labelled by @xmath78 or colored by @xmath276 since @xmath277 is simple . in the second case
, the third vertex of @xmath74 is @xmath266 which is labelled by @xmath78 or colored by @xmath278 @xmath215 we proceed by induction on @xmath50when @xmath279 the assertion is true since @xmath203 has only one face .
suppose that ( [ ins ] ) holds for @xmath62 and consider @xmath280write @xmath281 where @xmath282 .
then we have by induction @xmath283 thus it suffices to show that @xmath284 which immediately follows from the definition of @xmath18 and the description of the insertion algorithm given above
. * remark . *
binary trees and triangulations are known to be formally equivalent combinatorial objects .
this analogy can be extended to binary search trees which corresponds to simple colored triangulations .
the previous insertion scheme can be then regarded as an analogue of knuth s insertion algorithm on binary search trees @xcite .
fig10.eps + figure 6 : @xmath203 with @xmath285 and @xmath286[fig10 ] in this paragraph we prove that the above map @xmath288 is in fact a surjective _ morphism _ from the cayley graph of @xmath289 @xmath290 to the flip graph @xmath291 here @xmath292 is the graph on the vertex set @xmath1 , with edges all pairs @xmath293 such that @xmath294 differ by an elementary transposition . since we have identified the permutations of @xmath295 with the standard words of length @xmath5 with letters in @xmath115 this is equivalent to say that @xmath296 is obtained by switching two consecutive letters of @xmath0 , that is @xmath30{lcl}% \sigma & = & uxzv\\ \sigma^{\prime } & = & uzxv\label{edge}% \end{array}\ ] ] where @xmath297 are letters in @xmath77 and @xmath32 standard words on @xmath117 by symmetry , we can suppose @xmath298 .
[ lemfond ] let @xmath299 , where as above @xmath300 and @xmath32 are words on @xmath77 . 1 .
suppose that @xmath33 does not contain any letter @xmath232 such that @xmath301 .
then the associated triangulations @xmath13 and @xmath302 differ by a diagonal flip .
in other words , the edge @xmath303 in @xmath292 is mapped by @xmath17 to an edge in the flip graph @xmath291 2 .
suppose that @xmath33 contains a letter @xmath76 such that @xmath301 .
then @xmath304 here , the edge @xmath305 in @xmath292 is contracted by @xmath17 to a single vertex in @xmath291 @xmath204 since @xmath0 is a standard word , the letters @xmath76 such that @xmath301 , if any , belong to @xmath306set @xmath307then @xmath308set @xmath309 and @xmath310 then @xmath311 is obtained from @xmath312 by flipping the diagonal @xmath313 in the quadrilateral @xmath314thus @xmath13 and @xmath302 also differ by a diagonal flip in @xmath291 @xmath215 this times there exists a letter @xmath76 such that @xmath301 in @xmath315hence @xmath316 .
this implies immediately that @xmath317 and thus @xmath318 the map @xmath17 is a surjective morphism of graphs from @xmath319 to @xmath291 we have already obtained that @xmath17 is a surjective map .
now suppose that the permutation @xmath0 and @xmath296 belongs to the same edge in @xmath320then they can be written as in ( [ edge ] ) .
the previous lemma implies that we will either have @xmath321 or the pair @xmath322 differs by exactly one diagonal flip .
thus @xmath17 is a morphism of graphs , as claimed . in this section
we shall determine the fibers of the morphism @xmath323 .
it turns out that they can be characterized in terms of the _ sylvester relations
_ introduced by hivert , novelli and thibon in the context of binary search trees @xcite .
two words @xmath28 and @xmath29 on a totally ordered alphabet @xmath324 are said to be _
sylvester adjacent _ if there exist three words @xmath325 and three letters @xmath326 such that @xmath30{rcl}% w_{1 } & = & uxzu^{\prime}yu^{\prime\prime}\\ w_{2 } & = & uzxu^{\prime}yu^{\prime\prime}. \label{sylrel}% \end{array}\ ] ] two words @xmath4 and @xmath327 are _ sylvester congruent _ if there exists a chain of words @xmath328 such that for any @xmath329 the words @xmath330 and @xmath331 are sylvester adjacent . the sylvester congruence is a congruence denoted by @xmath332 on the free monoid @xmath333 on @xmath324 . we write @xmath334 $ ] for the sylvester class of the word @xmath335 the sylvester congruence has the following remarkable property .
[ lem - stand ] 1 .
let @xmath28 and @xmath29 two words on @xmath324 such that @xmath336then we have @xmath337 2 .
conversely if @xmath28 and @xmath29 have the same evaluation @xmath42 and verify @xmath338 we have @xmath339 the statement follows immediately from the definition of the standardization map and the definition ( [ sylrel ] ) of sylvester adjacency .
we now turn to the description of the fibers of @xmath288 in terms of the sylvester relations .
[ prop_sylvsn]consider @xmath340 and @xmath341 two permutations in @xmath119then @xmath342 in other words , two permutations are sylvester congruent if and only if they are readings of the same triangulation .
the proposition is immediate for @xmath343 , and so we assume now @xmath344 .
suppose that @xmath345by induction , it is enough to prove that @xmath346 when @xmath340 and @xmath341 are sylvester adjacent .
this directly follows from @xmath215 of lemma [ lemfond ] .
this shows in particular that for any triangulation @xmath347 the fiber @xmath148 is a nonempty disjoint union of sylvester classes . to obtain the left to right part of the proposition ,
observe first that all the words belonging to the same sylvester class have the same last letter . by an easy induction , this implies that each sylvester class contains words @xmath4 which verify the separation property @xmath348 where @xmath76 is a letter and @xmath32 are words either empty or containing letters greater than @xmath76 and smaller than @xmath76 , respectively .
such a word @xmath4 is not unique . to obtain a normal form it suffices to impose that @xmath349 and @xmath33 are empty or verify themselves the separation property .
denote by @xmath350 the number of such normal forms for the permutations of @xmath351then the separation property ( [ sp ] ) gives the following recurrence formula for the numbers @xmath350:@xmath352 moreover we have @xmath353 and @xmath354 .
thus the above recurrence formula is the recurrence formula for the catalan numbers @xmath355 .
this shows that the number of sylvester classes is equal to @xmath356 since @xmath17 is surjective , there are exactly @xmath355 nonempty fibers @xmath148 which all contains at least a sylvester class .
this means that each fiber is a sylvester class and immediately yields the desired implication @xmath357 the previous result easily extends to the setting of colored triangulations .
[ th_sylvt]consider @xmath186 such that @xmath358 and @xmath359 two words of evaluation @xmath42 on the alphabet @xmath15 .
then @xmath360 two words on @xmath15 are sylvester congruent if and only if they are readings of the same simple colored triangulation .
since @xmath28 and @xmath29 have the same evaluation @xmath361 we deduce from lemma [ lem - stand ] and proposition [ prop_sylvsn ] the equivalencies : @xmath362 * remarks . * @xmath98 the interpretation of the sylvester congruence we use in this paper is not that originally given in @xcite where the sylvester classes are defined as the fibers of the map which associated to each word the binary search tree obtained via knuth s insertion algorithm on binary trees .
this means that we have chosen to parametrize the sylvester classes by simple colored triangulations rather than binary search trees . to obtain the binary tree corresponding to a simple signed triangulation of canonical reading @xmath4 ,
one applies the insertion algorithm on binary trees starting from @xmath10conversely , the triangulation associated to a binary tree with right to left postfix reading @xmath4 , is the triangulation @xmath203 . note that @xmath4 is then the canonical reading of @xmath203 .
@xmath99 it is very easy to obtain the sylvester class of a word @xmath4 from its associated colored triangulation @xmath235indeed , @xmath334 $ ] is simply the set of readings of @xmath203 defined in [ suubsec_deffi ] .
@xmath363 for any simple triangulation @xmath23 we know by theorem [ th_sylvt ] that the set @xmath364 $ ] of all the different readings of @xmath27 is a plactic class . the same property holds for the set of readings @xmath25 $ ] of @xmath147then the map @xmath365 is a one - to - one correspondence between @xmath364 $ ] and @xmath25 $ ] .
[ prop_flip]two triangulations @xmath108 and @xmath109 differ by a diagonal flip if and only if there exist a reading @xmath28 of @xmath108 and a reading @xmath29 of @xmath109 such that @xmath30{rcl}% w_{1 } & = & uxzv\\ w_{2 } & = & uzxv \label{cond}% \end{array}\ ] ] where @xmath366 belong to @xmath77 and where @xmath367 are words such that @xmath33 contains no letter @xmath232 satisfying @xmath301 or @xmath368 if @xmath28 and @xmath29 are readings respectively of @xmath108 and @xmath109 verifying ( [ cond ] ) , they differ by a diagonal flip by @xmath204 of lemma [ lemfond ] .
conversely suppose that @xmath108 and @xmath109 differ by a diagonal flip .
denote by @xmath369 the quadrilateral in which the diagonal flip happens .
one can suppose that @xmath370 is the clockwise reading of the vertices of @xmath371 and consider that the diagonal @xmath313 is flipped in @xmath108 to give the diagonal @xmath372 in @xmath373 consider the triangulations @xmath374 and @xmath375 whose vertices are respectively indexed by the integers @xmath376 and @xmath377 in @xmath51 such that @xmath378 in @xmath108 and @xmath379 the vertices @xmath380 and @xmath256 are connected to edges which are not edges of @xmath381we have the same property for the vertices @xmath382 and @xmath256 in @xmath383 , and for the vectors @xmath382 and @xmath384 in @xmath385thus there is at least a reading of @xmath108 and @xmath109 which starts with the vertices of @xmath386 distinct of @xmath387 and @xmath388choose one of these readings and denote by @xmath349 the word obtained .
then @xmath349 contains all the vertices labelled by an integer @xmath76 such that @xmath389 moreover , we can go on the lecture of @xmath108 by reading successively @xmath297 and next choose a reading @xmath33 of the remaining triangulation .
similarly , we can go on the lecture of @xmath109 by reading successively the vertices @xmath390 and next form the word @xmath33 as previously .
finally the readings we obtain for @xmath108 and @xmath109 are respectively @xmath391 and @xmath392 and the integers @xmath76 such that @xmath301 are in @xmath393 * remark . *
this proof shows also that the letters @xmath256 and @xmath382 appearing in ( [ cond ] ) are the labels of the faces of the quadrilateral of @xmath108 in which the flip happens .
it is well known that , considered as a statement of graph theory , the four color theorem is equivalent to say that every simple finite planar graph admits a proper four coloring of its vertices . in @xcite ,
the first author obtains a reformulation of this theorem in terms of signed paths between triangulations of polygons .
let us recall briefly the main ideas which permit this reformulation .
tutte has proved in @xcite that every @xmath45-connected finite planar graph is hamiltonian ( i.e. admits a cycle which visits each vertex exactly once ) .
this implies that it suffices to prove the four color theorem for hamiltonian planar triangulations of the sphere @xmath394 .
for such a triangulation @xmath395 , it is natural to confine the hamiltonian path @xmath14 on the equator of @xmath394 .
the path @xmath14 can be then regarded as a polygon and one defines two triangulations @xmath108 and @xmath109 of @xmath14 by considering the sub - graphs of @xmath395 lying respectively in the northern and southern hemispheres of @xmath394 .
conversely , two triangulations @xmath108 and @xmath109 of the same polygon @xmath14 define a triangulation @xmath396 of @xmath394 obtained by embedding them in the two hemispheres of @xmath394 and gluing them along their common boundary @xmath14 confined on the equator .
we denote by @xmath397 the set of faces of @xmath395 and by @xmath398 the subset of faces incident to some vertex @xmath33 .
a signing of @xmath395 is a map @xmath399 which associates to each face of @xmath395 one of the integers @xmath400 or @xmath401we then denote by @xmath402 the signed triangulation obtained . given @xmath33 a vertex of @xmath403 write@xmath404 for the sum of signs of the faces @xmath74 incident to @xmath68the signing @xmath24 is a _ heawood signing _ if at each vertex @xmath33 of @xmath405 one has@xmath406 we will then say that @xmath403 is a heawood signed triangulation . at the end of the @xmath407-th century
, heawood @xcite has proved that @xmath395 admits a proper four coloring of its vertices if and only if there exists a heawood signing on its faces .
now consider a triangulation @xmath21 of the polygon @xmath36as observed in @xcite there is a very simple way to obtain from @xmath21 a triangulation of the sphere which admits a heawood signing . indeed , for @xmath408 every northern face has a corresponding southern face having the same three vertices .
so its suffices to define @xmath24 such that these faces have opposite signs to obtain a heawood signing .
the diagonal flip operations on the triangulation @xmath395 are natural geometrical transformations which yields new triangulations from @xmath395 .
two adjacent faces @xmath74 and @xmath409 in the same plane trinagulation @xmath108 or @xmath109 defining @xmath395 form a quadrilateral @xmath371 and the bound between @xmath74 and @xmath409 coincide with a diagonal @xmath410 of @xmath411the diagonal flip operation in @xmath371 delete the diagonal @xmath410 and replace it by the opposite diagonal of @xmath412 to obtain diagonal flips operations on heawood signed triangulations ( that is which preserve the heawood property ) , one has to restrict the authorized diagonals flips to what we call _ signed flips _ , defined as flips for which the signs of the two faces of @xmath371 are the same and are changed into their opposite during the flip operation . consider @xmath403 a hamiltonian planar triangulation of the sphere and denote by @xmath413 and @xmath414 the two signed plane triangulations such that @xmath415 .
suppose that there exists a sequence of signed flips from @xmath413 to @xmath416since @xmath417 is a heawood signed triangulation and by using that the signed flips preserve the heawood property , one then obtains that @xmath403 is a heawood triangulation of the sphere .
thus the existence of a signed flip sequence between any two triangulations of a polygon implies the four color theorem .
this is the result obtained in @xcite .
the converse is true as proved by gravier and payan in @xcite .
consider a triangulation @xmath21 of a convex @xmath2-gon @xmath14 .
for any @xmath418 , denote by @xmath27 the signed triangulation obtaining by signing @xmath21 following @xmath89note that @xmath27 is not simple in general .
let @xmath419 be the set of all signed triangulations obtained by applying a sequence of signed flips starting from @xmath27 .
for any signed triangulation @xmath420 belonging to @xmath419 , we will say that there exists a _ signed path _ between @xmath27 and @xmath420 .
[ lem - esse]suppose that @xmath419 contains signed triangulations @xmath420 and @xmath421 with the same underlying triangulation @xmath422 . then @xmath423 .
since the signed sphere triangulation @xmath424 has the heawood property , and since signed flips preserve this property , it follows that the signed sphere triangulation @xmath425 also has the heawood property .
we now deduce from this that @xmath426 by induction on the number @xmath61 of vertices .
the statement is trivial for @xmath134 , as there is only one face in @xmath422 .
assume @xmath427 , and let @xmath33 be an ear in @xmath422 .
thus @xmath33 is contained in a unique face @xmath74 of @xmath422 , and therefore is contained in exactly two faces of @xmath428 , namely one copy of @xmath74 on each hemisphere .
since the signs of these two faces must sum up to 0 mod 3 in @xmath429 by the heawood property , it follows that @xmath430 .
let @xmath371 denote the polygon obtained by contracting one of the two edges of @xmath14 containing @xmath33 , let @xmath431 denote the triangulation of @xmath371 obtained by cutting the ear @xmath33 in @xmath422 , and let @xmath432 , respectively @xmath433 , denote the restrictions of @xmath434 and @xmath435 to the faces of @xmath431 .
then the signed sphere triangulation @xmath436 still has the heawood property , as easily seen .
it follows by the induction hypothesis that @xmath437 .
therefore , @xmath423 as claimed . * remark .
* by the previous lemma , there exists a signed path between the signed triangulations @xmath27 and @xmath438 only if there exists a path without loop between their underlying triangulations @xmath21 and @xmath422 in @xmath287 .
write @xmath439 for the set of triangulations obtained by deleting the signs @xmath440 and @xmath441 in the signed triangulations of @xmath442we deduce from @xcite and @xcite the following reformulation of the four color theorem .
[ ref1]the four color theorem is equivalent to the following statement .
for any triangulation @xmath102 , we have @xmath443 in other words , for any pair @xmath444 of triangulations in @xmath63 , there exist @xmath445 and a sequence of signed flips from @xmath27 to @xmath446 .
representing triangulations by permutations via the map @xmath447 , and using the preceding theorem , we obtain in this section a reformulation of the four color theorem using now signed permutations .
consider the alphabet @xmath448 and set @xmath449 @xmath450the letters of @xmath451 ( resp .
@xmath452 ) are said _ negative _ ( resp . _ positive _ ) .
define the bar involution on the letters @xmath34 of @xmath453 by @xmath454 write @xmath455 if @xmath456 and @xmath457 if @xmath458for any word @xmath459 on @xmath460 set @xmath461the set of signed permutations on @xmath453 is defined by @xmath462 two words @xmath28 and @xmath29 of @xmath463 differ by an authorized transposition if one of the two situations happens : 1 .
@xmath464 and @xmath465 are sylvester adjacent , that is there exist letters @xmath466 such that @xmath467 and @xmath30{rcl}% w_{1 } & = & u\ \alpha\ \gamma\ v\ \beta\ w\\ w_{2 } & = & u\ \gamma\ \alpha\ v\ \beta\ w \label{k1}% \end{array}\ ] ] where @xmath468 are factors of @xmath28 .
2 . there exist letters @xmath31 with the same sign such that @xmath30{rcl}% w_{1 } & = & u\ \alpha\ \gamma\ v\\ w_{2 } & = & u\ \overline{\gamma}\ \overline{\alpha}\ v\label{k2}% \end{array}\ ] ] where @xmath32 are factors of @xmath28 and @xmath33 does not contain any letter @xmath34 such that @xmath469 ( i.e. @xmath464 and @xmath465 are not sylvester adjacent ) .
consider @xmath0 and @xmath296 two permutations of @xmath119we will say that there exists a signed path between @xmath0 and @xmath296 if one can find two words @xmath470 in @xmath463 such that @xmath471 @xmath472 and @xmath327 can be obtained from @xmath4 by applying successive authorized transpositions . from proposition [ prop_sylvsn ] and theorem [ ref1 ]
we deduce the following reformulation of the four color theorem : [ fc]the four color theorem is equivalent to the following statement : for any positive integer @xmath5 , there exists at least a signable path joining two permutations of @xmath1 .
to each signed triangulation @xmath27 with @xmath473 we associate the set of signed permutations @xmath474_{s}=\{w=\beta_{1}\cdot\cdot\cdot\beta_{n}\mid\left|
w\right| \in\lbrack t]\text { and for each } \diamondsuit\in\{-,+\},\beta _ { i}\in i_{\diamondsuit}\longleftrightarrow\varepsilon_{i}=\diamondsuit\}\ ] ] that is , @xmath364_{s}$ ] is the set of signed permutations obtained by signing the words contained in the sylvester class of @xmath21 so that the sign associated to each @xmath82 is equal to @xmath475 by proposition [ prop_sylvsn ] , two signed permutations @xmath327 and @xmath4 belong to @xmath364_{s}$ ] if and
only if they differ by successive transpositions of kind ( [ k1 ] ) . given two triangulations @xmath21 and @xmath107 in @xmath476 we said that there exists a signable flip path between @xmath21 and @xmath107 if @xmath107 belongs to @xmath477 ( with the notation of [ subsec_fourcol ] ) .
we deduce from proposition [ prop_flip ] that the signed triangulations @xmath27 and @xmath478 differ by a signed flip if and only if there exist a signed word in @xmath364_{s}$ ] and a signed word in @xmath479_{s}$ ] which differ by a transposition of kind ( [ k2 ] ) .
thus , @xmath27 and @xmath480 differ by a signed flip if and only if each signed word of @xmath479_{s}$ ] can be obtained by applying successive transpositions of kind ( [ k1 ] ) or ( [ k2 ] ) from any signed word of @xmath364_{s}.$ ] by transitivity , we deduce that the existence of a signable flip path between @xmath21 and @xmath107 is equivalent to that of a signable path between each reading of @xmath21 and each reading of @xmath481 now , by theorem [ ref1 ] the four color theorem is equivalent to the statement : for any positive integer @xmath5 , there exists at least a signable flip path joining two triangulations of @xmath63.by the previous argument , this implies our theorem .
* example . *
consider @xmath30{rcl}% \sigma_{1 } & = & 324156,\\ \sigma_{2 } & = & 453126 .
\end{array}\ ] ] here is a signed path between suitable signings of @xmath340 and @xmath341 , where @xmath482 indicates a signed flip and @xmath332 a sylvester adjacency : @xmath483 this signed path produces an explicit heawood signing , and hence an explicit proper four vertex - coloring , of the sphere triangulation obtained by gluing the octogon triangulations @xmath484 and @xmath485 along their boundary
. * remark .
* there exists a simple procedure deciding whether a given path @xmath486 in @xmath292 is signable .
the path @xmath486 is signable if one can compute a sequence @xmath487 of signed permutations such that @xmath488 for any @xmath489 by the following procedure .
first , the permutations @xmath340 and @xmath341 differ by a transposition , thus can be written @xmath30{rcl}% \sigma_{1 } & = & u\ a\ c\ v\\ \sigma_{2 } & = & u\ c\ a\ v. \end{array}\ ] ] where @xmath490 . if @xmath340 and @xmath341 differ by a sylvester relation , set @xmath491otherwise set @xmath492 .
now suppose we have obtained the signed permutations @xmath493 from @xmath494 .
since @xmath495 and @xmath496 differ by a transposition we have@xmath30{lcl}% w_{i } & = & u_{i}\ \alpha_{i}\ \gamma_{i}\ v_{i}\\ \sigma_{i+1 } & = & u_{i}\ c_{i}\ a_{i}\ v_{i}% \end{array}\ ] ] where @xmath497 and @xmath498 .
if @xmath495 and @xmath496 differ by a sylvester relation , set @xmath499otherwise two situations can happen .
* when @xmath500 and @xmath501 have opposite signs the algorithm stops and the path is not signable . * when @xmath500 and @xmath501 have the same sign , set @xmath502 . * further remarks .
* @xmath98 to decide if there exists a signed path between the two permutations @xmath0 and @xmath503 it is sufficient by lemma [ lem - esse ] to test the paths joining these vertices in @xmath319 by restricting to the paths with no loop .
@xmath99 if @xmath486 is signable in @xmath290 then @xmath504 where for any @xmath505 @xmath506 $ ] is a signable path in @xmath287 .
we are going to describe an alternative formulation of this theorem using signings of the _ diagonals _ in the triangulations , rather than of the faces , to be used later in subsection [ subsec_asso ] .
consider a signed triangulation @xmath27 and @xmath71 a diagonal of @xmath147let @xmath507 and @xmath508 be the signs of the two faces of @xmath27 adjacent to @xmath509 then label the diagonal @xmath71 by the product @xmath510by proceeding similarly for all the diagonals of @xmath347 we obtain a triangulation with signed diagonals @xmath511 .
note that the signing @xmath511 of the diagonals of @xmath21 determines the signing of its faces up to an overall inversion of signs .
indeed , the signing of the faces of @xmath21 can be obtained from the sign of one face and the signs of the diagonals of @xmath21 .
suppose that the diagonal @xmath71 in @xmath21 is flipped in @xmath512 and write @xmath371 for the quadrilateral whose diagonals are @xmath71 and @xmath512 .
the signed flip of the diagonal @xmath71 is defined in @xmath511 when the sign attached to @xmath71 is @xmath513 in this case , this yields a triangulation with signed diagonals defined by the three followings requirements : 1 .
@xmath512 is signed by @xmath441 2 .
the signs of the diagonals of @xmath511 which are edges of @xmath371 are changed into their opposite 3 .
the signs of the remaining diagonals are unchanged .
note that the signed flip is not defined when the sign of @xmath71 is @xmath440 .
denote by @xmath514 the subgraph of the flip graph generated from @xmath511 by applying signed flips and write @xmath515 for the set of triangulations obtained by deleting the signs @xmath440 and @xmath441 in the triangulations with signed diagonals belonging to @xmath516by definition of the signed flips on triangulations with signed diagonals , we have @xmath517 and theorem [ ref1 ] is equivalent to the assertion : for any triangulation @xmath102@xmath518 where @xmath519 yields a signing @xmath511 of the @xmath62 diagonals of @xmath21 .
we suppose in the sequel that we have chosen a labelling of the @xmath520 diagonals of @xmath14 ( for example we can consider the labelling by roots belonging to the root system of type @xmath521 depicted in @xcite ) .
consider a path @xmath522 in @xmath287 where for @xmath523 the triangulations @xmath86 and @xmath524 differs by a diagonal flip .
the path @xmath525 is said signable if for each @xmath526 there exists a signing @xmath527 of the diagonals in @xmath86 such that the transformation @xmath528 is a signed flip ( that is , conditions @xmath529 and @xmath530 above are verified ) .
in fact it is rather easy to determinate if a path is signable or not . for any @xmath531 write @xmath532 for the set of diagonals in @xmath86 which has appeared after a flip @xmath533 @xmath534 .
this means that @xmath532 is the set of diagonals of @xmath86 which result of a flip operation at one step of the path @xmath535 .
consider first the flip @xmath536 of the diagonal @xmath71 into @xmath512 .
we have @xmath537 and the sign of @xmath512 in @xmath109 must be @xmath513now suppose by induction that we have determined the signs of the diagonals of @xmath532 @xmath538 so that at each step @xmath539 @xmath540 the signs obtained are compatible with a signed flip ( that is with conditions @xmath529 and @xmath530 above ) .
let @xmath541 be the diagonal flipped in @xmath86 and @xmath542 the new diagonal obtained in @xmath543when @xmath541 has sign @xmath544 the path is not signable .
when @xmath541 is not signed or is signed by @xmath545 we have @xmath546 . in @xmath547 we sign @xmath542 with @xmath441 and we change the signs of the edges of the quadrilateral associated to @xmath541 and @xmath542 which belong to @xmath548 the path @xmath525 is signable if the previous algorithm does not stop until the last flip @xmath549 has been considered . in this case , it becomes immediate to complete in each @xmath86 the signs of the diagonals which are not in @xmath548 indeed , its suffices to sign arbitrary the unsigned diagonals of @xmath550.@xmath227the complete signings in @xmath551 are then determined by conditions @xmath529 and @xmath530 above .
figure11.eps + figure 7 : signable and nonsignable paths in @xmath287 [ fig11 ] given a path @xmath552 we denote by @xmath553 the subset of diagonals on @xmath14 which appear in the triangulations @xmath554 .
we endow @xmath555 with the structure of an oriented graph by drawing an edge @xmath556 between the diagonals @xmath71 and @xmath512 if there exists @xmath557 such that @xmath524 is obtained by flipping @xmath558 into @xmath512
. * remarks . * @xmath98 if @xmath522 is a signable path , the paths @xmath559 @xmath560 are also signable .
@xmath99 if all the chains in @xmath555 have length less than or equal to @xmath43 ( i.e. each diagonal is flipped at most on time ) , the path @xmath522 is signable . indeed @xmath561 or for any @xmath562 the diagonal @xmath541 in the above procedure is never signed .
@xmath363 suppose that @xmath21 and @xmath422 are two triangulations of @xmath287 . by lemma [ lem - esse ]
the problem of determining whether there exists a signable diagonal path between @xmath21 and @xmath422 can be solved by considering only paths without loop between @xmath21 and @xmath422 in @xmath563 .
there is a finite number of such paths and one can apply to each of them the previous procedure .
> from the above arguments and theorem [ ref1 ] we obtain the following reformulation of the four color theorem : [ cor](of theorem [ ref1 ] ) the four color theorem is equivalent to the following statement : for any positive integer @xmath165 there exists at least a signable path between two triangulations of the @xmath2-gon .
the flip graph @xmath287 is the @xmath400-skeleton of a convex polytope called the @xmath5-dimensional associahedron @xmath564 ( see @xcite ) .
the vertices of the associahedron can be identified with the triangulations of the @xmath2-gon @xmath14 and its facets with the @xmath565 diagonals of @xmath36its edges correspond to partial triangulations of @xmath14 .
given a star @xmath566 write @xmath567 for the vertex of maximal order in @xmath568 .
this yields a natural one - to - one correspondence between the stars and the vertices of @xmath36the vertices of @xmath564 belonging to the facet @xmath569 corresponding to the diagonal @xmath71 coincide with the triangulations which contains @xmath71 .
this implies that each facet @xmath569 contains exactly two stars .
more precisely we have : [ lem_facet]let @xmath570 and @xmath571 be two stars considered as vertices of @xmath564
. then we have the following equivalences : 1 .
@xmath570 and @xmath571 belong to the same facet @xmath569 if and only if @xmath572 2 .
@xmath570 and @xmath571 do not belong to the same facet @xmath569 if and only if @xmath573 are two consecutive vertices of @xmath14 ( that is form an edge of @xmath14 ) . consider a triangulation @xmath21 and a diagonal @xmath71 in @xmath21 .
denote by @xmath107 the triangulation obtained by flipping @xmath71 in @xmath21 and by @xmath512 the diagonal of @xmath107 such that @xmath574 .
the diagonal flip @xmath575 can be interpreted as a move from @xmath569 to @xmath576 in @xmath564 along the edge defined by the partial triangulation @xmath577.@xmath578 in this case , the facets @xmath569 and @xmath576 are disjoint .
indeed the diagonals @xmath71 and @xmath512 are secant in @xmath110 thus one can not find a triangulation @xmath21 belonging to @xmath569 and @xmath576 .
an edge @xmath579 of @xmath564 is contained in @xmath580 facets .
suppose that @xmath581 where @xmath21 and @xmath107 are triangulations belonging respectively to @xmath569 and @xmath582then @xmath107 is obtained from @xmath21 by flipping @xmath71 into @xmath583denote by @xmath371 the quadrilateral in @xmath14 whose diagonals are @xmath71 and @xmath583the edges of @xmath371 are either diagonals either edges of @xmath14 .
moreover , for @xmath584 the number of edges of @xmath371 which are diagonals of @xmath14 belongs to @xmath585to distinguish the facets corresponding to diagonals of @xmath14 among the facets containing @xmath586 it suffices to consider the four stars @xmath587 and @xmath588 which appear respectively in @xmath569 and @xmath582then the facets corresponding to diagonals of @xmath14 are the diagonals which can be obtained by connecting two vertices @xmath589 @xmath590by lemma [ lem_facet ] , this is equivalent to find the pairs of vertices @xmath591 in @xmath564 which belongs to the same facet .
the pairs @xmath592 and @xmath593 correspond respectively to @xmath71 and @xmath512 . in the sequel we will denote by @xmath594 the set of facets corresponding to the edges of @xmath371 which are diagonals of @xmath14 distinct of @xmath71 and @xmath512 . by the previous arguments @xmath594 coincide with the facets containing a pair @xmath591 distinct of @xmath592 and @xmath593 where @xmath595 are the stars appearing in @xmath569 and @xmath576 the two facets of @xmath564 connected by @xmath579 .
our aim is now to obtain a reformulation of the four color theorem using only the geometry of the associahedron and the distinguish subset @xmath596 of its vertices which correspond to stars .
a walk @xmath597 of length @xmath598 on @xmath564 is determined by @xmath598 vertices @xmath599 such that for any @xmath523 @xmath600 is an edge of @xmath601 in this case write @xmath602 for each edge @xmath600 , denote by @xmath603 @xmath604 the unique pair of facets in @xmath564 such that @xmath605 @xmath606 and @xmath607set @xmath608 @xmath609 note that @xmath610 in general ( see example below ) .
the walk @xmath597 is signable if the facets of @xmath611 can be signed by the following recursive procedure .
first sign the face @xmath612 with @xmath513suppose that the facets of @xmath613 are signed . if @xmath614 is signed by a @xmath544 the algorithm stops . otherwise , consider the edge @xmath600 and the set of facets @xmath615 . then sign the face @xmath616 with a @xmath441 and change the sign of the facets of @xmath617 > from corollary [ cor ] and the above arguments
, we derive the following reformulation of the four color theorem in terms of the geometry of the associahedron : the four color theorem is equivalent to the following statement : for any positive integer @xmath165 there is a least a signable walk between two vertices of the associahedron @xmath564 .
* remarks .
* @xmath98 by lemma [ lem - esse ] , the problem of finding a signed walk between two vertices @xmath431 and @xmath618 of the associahedron can be solved by applying the previous procedure to the walks joining @xmath431 to @xmath618 in which each vertex is attained at more one time ( that is by excluding the walks with loops ) .
@xmath99 at each step of the above procedure , the set @xmath615 is determined only by the vertices of @xmath619 and @xmath620 which belong to @xmath596 .
the signable path of figure 7 is equivalent to the signed walk on @xmath621 given in the figure below@xmath622{10.2692cm}{\begin{center } \includegraphics [ height=12.1166 cm , width=10.2692 cm ] % { figure12.eps}% \\ figure 8 \end{center}}}% % endexpansion\ ] ] where the solid ( resp .
dashed ) signs belong to apparent ( resp .
non apparent ) facets .
the set @xmath63 can be identified with the set of simple colored triangulations @xmath623 with @xmath624 .
the flip graph can be generated starting from any triangulation , by applying diagonal flips .
this yields to the following natural problem : consider a colored triangulation @xmath27 and denote by @xmath625 the subgraph of the flip graph generated from @xmath27 by applying homogeneous flips .
the coloring @xmath24 defines subtriangulations in @xmath27 obtained by gluing together the adjacent faces which have the same color .
these subtriangulations will be called the connected components of @xmath626denote them by @xmath554 . for any @xmath505 let @xmath627 be the number of faces in @xmath628
since the homogeneous flips stabilize the connected components @xmath86 , we obtain : the above isomorphism can be explicited by associating to each colored triangulation of @xmath625 the @xmath598-uple of triangulations defined from its connected components @xmath554 as pictured in the figure below .
this answers problem [ problem ] .
consider @xmath630 and @xmath249 a simple colored triangulation of @xmath631.@xmath227denote by @xmath632 the subgraph of the flip graph generated from @xmath27 by applying switched flips .
write @xmath633 for the graph obtained by drawing an edge between two simple triangulations of @xmath634 when they differ by a switched flip .
we are going to show that @xmath635 is connected this will imply the equality @xmath636 for any simple colored triangulation such that @xmath637 .
let @xmath39 be the frobenius subgroup of @xmath1 defined by @xmath361 that is the subgroup of permutations which stabilize the intervals @xmath638 and @xmath639 for @xmath640 the elements of the coset @xmath41 will be identified with the words of length @xmath5 and evaluation @xmath42 on the totally ordered alphabet @xmath641 that is we set @xmath642
. denote by @xmath643 the cayley graph of @xmath644 this means that the vertices of @xmath155 are the words of length @xmath5 and evaluation @xmath42 and there is an edge between @xmath4 and @xmath327 if and only if @xmath327 is obtained by switching two adjacent letters of @xmath645 denote by @xmath646 the standardization map on simple colored triangulations defined by @xmath647 from the definition of the standardization map and since switched flips are particular cases of flips , we have : 1
. the standardization map @xmath365 on words is an injective morphism of graphs from @xmath643 to @xmath319 .
the standardization map @xmath646 is an injective morphism of graphs from @xmath648 to @xmath287 .
[ sflip]let @xmath249 and @xmath649 two triangulations in @xmath648
. then @xmath249 and @xmath649 differ by a switched flip if and only if there exist a reading @xmath4 of @xmath650 and a reading @xmath327 of @xmath649 of the form @xmath30{lcl}% w & = & uxzv\\ w^{\prime } & = & uzxv \label{sw}% \end{array}\ ] ] where @xmath297 are letters of @xmath15 and @xmath32 words on @xmath15 such that @xmath33 does not contain any letter @xmath76 verifying @xmath651 or @xmath652 .
by symmetry we only consider the case @xmath651 .
suppose that @xmath4 and @xmath327 are readings of @xmath249 and @xmath653 verifying @xmath654 and @xmath655 as in the theorem .
then @xmath656 and @xmath657 where @xmath658 are letters of @xmath77 and @xmath659 words on @xmath77 such that @xmath660 does not contain any letter @xmath76 verifying @xmath389 thus by applying @xmath400 of proposition [ prop_flip ] , @xmath21 and @xmath107 differs by a diagonal flip and the faces corresponding to its flip are labelled by @xmath661 and @xmath662 this implies that @xmath249 and @xmath663 differ by a flip corresponding to faces colored by @xmath256 and @xmath382 , hence by a switched flip .
conversely , suppose that @xmath249 and @xmath664 differ by a switched flip .
then @xmath21 and @xmath107 differ by a flip and by @xmath43 of proposition [ prop_flip ] , we have readings @xmath665 and @xmath666 respectively of @xmath21 and @xmath107 where @xmath667 are letters of @xmath77 and @xmath659 words on @xmath77 such that @xmath660 does not contain any letter @xmath76 verifying @xmath389 hence by applying the destandardization procedure ( see [ subsec_asso ] ) to @xmath0 and @xmath296 , one can find readings @xmath654 and @xmath655 respectively of @xmath249 and @xmath649 with @xmath668 1 .
the following diagram commutes:@xmath30{ccc}% \mathrm{cay}(\mathcal{c}_{n,\mu } ) & \overset{\mathrm{std}}{\rightarrow } & \mathrm{cay}(s_{n})\\ \phi\downarrow & & \downarrow\varphi\\ \mathcal{sf}_{n}(\varepsilon_{\mu } ) & \underset{\mathrm{std}}{\rightarrow } & \mathcal{f}_{n}% \end{array}\ ] ] where @xmath669 and @xmath646 are morphisms of graphs . 2 .
@xmath648 is connected .
@xmath670 we have already seen that @xmath18 is a surjective map .
suppose that @xmath4 and @xmath327 are words in @xmath643 which differ by the transposition of two consecutive letters .
if @xmath4 and @xmath327 are sylvester adjacent , @xmath157 and @xmath671 are also sylvester adjacent , thus @xmath672 .
this implies that @xmath673 since @xmath4 and @xmath327 have the same evaluation .
if @xmath4 and @xmath327 are not sylvester adjacent , they verify ( [ sw ] ) and by proposition [ sflip ] , we know that @xmath203 and @xmath674 differ by a switched flip in @xmath648 .
this proves that @xmath18 is a morphism of graphs .
now for any word @xmath4 in @xmath643 , @xmath675 by definition of the map @xmath676thus by lemma [ lemstd ] , the above diagram is a commutting diagram of morphisms
. * remark .
* the theorem implies in particular that @xmath678 for any simple colored triangulation @xmath679 .
thus it answers to problem [ problem ] when the colored triangulation @xmath27 is simple .
when @xmath27 is not simple , we have find no algebraic interpretation of the graph @xmath680 | to each permutation @xmath0 in @xmath1 we associate a triangulation of a fixed @xmath2-gon .
we then determine the fibers of this association and show that they coincide with the sylvester classes depicted in @xcite .
a signed version of this construction allows us to reformulate the four color theorem in terms of the existence of a signable path between any two permutations in the cayley graph of the symmetric group @xmath3 5.5 mm |
the block cipher algorithms are a family of cipher algorithms which use symmetric key and work on fixed length blocks of data .
since novembre 26 , 2001 , the block cipher algorithm `` rijndael '' , became the successor of des under the name of `` advanced encryption standard '' ( aes ) .
its designers , joan daemen and vincent rijmen used algebraic tools to give to their algorithm an unequaled level of assurance against the standard statistical techniques of cryptanalysis .
the aes can process data blocks of 128 bits , using cipher keys with lengths of 128 , 192 , and 256 bits @xcite .
one of the major issues of cryptography is the cryptanalysis of cipher algorithms .
cryptanalysis is the study of methods for obtaining the meaning of encrypted information , without access to the secret information that is normally required .
some mechanisms for breaking codes include differential cryptanalysis , advanced statistics and brute - force .
recent works like @xcite , attempt to use algebraic tools to reduce the cryptanalysis of a block cipher algorithm to the resolution of a system of quadratic equations describing the ciphering structure . as an example ,
nicolas courtois and josef pieprzyk have described the aes-128 algorithm as a system of 8000 quadratic equations with 1600 variables @xcite .
unfortunately , these approaches are infeasible because of the difficulty of solving large systems of equations .
we will also use algebraic tools but in a new way by using boolean functions and their properties .
our aim is to describe a block cipher algorithm as a set of boolean functions then calculate their algebraic normal forms by using the mbius transforms . in our study
, we will test our approach on the aes algorithm .
our goal is to describe it under the form of systems of boolean functions and to calculate their algebraic normal forms by using the mbius transforms .
the system of equations obtained is more easily implementable and could open new ways to cryptanalysis of the aes .
let be the set @xmath3 and @xmath4 a boolean algebra , then @xmath5 such that @xmath6 and @xmath7 , is a subset of @xmath8 containing all @xmath2-tuples of @xmath9 and @xmath10 .
the variable @xmath11 is called boolean variable if she only accepts values from @xmath12 , that is to say , if and only if @xmath13 or @xmath14 regardless of @xmath7 .
a boolean function of degree @xmath2 with @xmath1 is a function @xmath15 defined from @xmath16 , that is to say built from boolean variables and agreeing to return values only in the set @xmath3 .
for example , the function @xmath17 defined from @xmath18 is a boolean function of degree two with : @xmath19 let @xmath2 and @xmath20 be two positive integers .
a vector boolean function is a boolean function @xmath15 defined from @xmath21 .
an s - box is a vector boolean function .
finally , we can define a random boolean function as a boolean function @xmath15 whose values are independent and identically distributed random variables , that is to say : @xmath22 = \frac{1}{2}\ ] ] the number of boolean functions is limited and depends on @xmath2 .
thus , there is @xmath23 boolean functions .
similarly , the number of vector boolean functions is limited and depends on @xmath2 and @xmath20 .
thus , there exists @xmath24 vector boolean functions .
if we take , for example , @xmath25 then there exists @xmath26 boolean functions of degree two .
these 16 boolean functions are presented in the table in figure [ fig:16bool ] page . among the boolean functions of degree 2 , the best known
are the functions or , and and xor ( see fig .
[ fig : or ] , page ) , ( see fig . [
fig : and ] , page ) and ( see fig . [
fig : xor ] , page ) . [ cols="^,^",options="header " , ] the weight of the function @xmath15 is @xmath27 .
so we can reduce @xmath15 to the sum of 3 atomic functions @xmath28 , @xmath29 and @xmath30 . the function @xmath31 if and only if @xmath32 , @xmath33 and @xmath34 . from this
we can deduce that the anf of the function @xmath28 can be obtained by expanding the product @xmath35 . applying this reasoning to the functions @xmath29 and
@xmath30 we get the following equation : @xmath36
after this brief presentation of boolean functions , we have the necessary tools for the development of systems of boolean equations describing the _ advanced encryption standard_. we have just seen how to generate normal algebraic form ( anf ) of a boolean function
. the presented method is not easily automatable in a computer program .
so we will prefer the use of the mbius transform .
the mbius transform of the boolean function @xmath15 is defined by @xcite : @xmath37 with @xmath38 if and only if @xmath39 .
from there , we can define the normal algebraic form of a boolean function @xmath15 in @xmath2 variables : @xmath40 to better understand the mechanisms involved in the use of the mbius transform , take an example with the ` majparmi3 ` .
this function from @xmath41 is characterized by the truth table shown in figure [ fig : majparmi3 ] page .
> p18pt | > p18pt | > p18pt | > p52pt @xmath42 & @xmath43 & @xmath44 & ` majparmi3 ` @xmath9 & @xmath9 & @xmath9 & @xmath9 @xmath9 & @xmath9 & @xmath10 & @xmath9 @xmath9 & @xmath10 & @xmath9 & @xmath9 @xmath9 & @xmath10 & @xmath10 & @xmath10 @xmath10 & @xmath9 & @xmath9 & @xmath9 @xmath10 & @xmath9 & @xmath10 & @xmath10 @xmath10 & @xmath10 & @xmath9 & @xmath10 @xmath10 & @xmath10 & @xmath10 & @xmath10 calculating the mbius transform of the function we get the result of figure [ fig : rm - mp3 ] page . > p10pt | > p10pt | > p10pt | > p52pt > p10pt > p10pt | > p10pt | > p10pt | > p10pt | > p30pt | @xmath42 & @xmath43 & @xmath44 & ` majparmi3 ` & @xmath45 & & @xmath46 @xmath9 & @xmath9 & @xmath9 & @xmath9 & @xmath45 & @xmath9 & @xmath9 & @xmath9 & @xmath9 & @xmath9 @xmath9 & @xmath9 & @xmath10 & @xmath9 & @xmath45 & @xmath9 & @xmath9 & @xmath9 & @xmath9 & @xmath9 @xmath9 & @xmath10 & @xmath9 & @xmath9 & @xmath45 & @xmath9 & @xmath9 & @xmath9 & @xmath9 & @xmath9 @xmath9 & @xmath10 & @xmath10 & @xmath10 & @xmath45 & @xmath10 & @xmath10 & @xmath10 & @xmath10 & @xmath10 @xmath10 & @xmath9 & @xmath9 & @xmath9 & @xmath45 & @xmath9 & @xmath9 & @xmath9 & @xmath9 & @xmath9 @xmath10 & @xmath9 & @xmath10 & @xmath10 & @xmath45 & @xmath10 & @xmath10 & @xmath10 & @xmath10 & @xmath10 @xmath10 & @xmath10 & @xmath9 & @xmath10 & @xmath45 & @xmath10 & @xmath10 & @xmath10 & @xmath10 & @xmath10 @xmath10 & @xmath10 & @xmath10 & @xmath10 & @xmath45 & @xmath10 & @xmath9 & @xmath10 & @xmath9 & @xmath9 after the mbius transform of the function obtained , we take the @xmath47 for which @xmath48 . in our case
we have the triplets @xmath49 , @xmath50 , @xmath51 from which we can deduce the equation : @xmath52 with the addition corresponding to a ` xor ` and multiplication to a ` and ` .
the implementation of the mbius transform in python is performed by the two functions described in the listing [ lst : tm ] page . .... def xortab(t1 , t2 ) : " " " takes two tabs t1 and t2 of same lengths and returns t1 xor t2 . " " " result = '' for i in xrange(len(t1 ) ) : result + = str(int(t1[i ] ) ^ int(t2[i ] ) ) return result def moebiustransform(tab ) : " " " takes a tab and return tab[0 : len(tab)/2 ] , tab[0 : len(tab)/2 ] ^ tab[len(tab)/2 : len(tab ) ] .
usage : moebiustransform(1010011101010100 ) -- > [ 1100101110001010 ] " " " if len(tab ) = = 1 : return tab else : t1 = tab[0 : len(tab)/2 ] t2 = tab[len(tab)/2 : len(tab ) ] t2 = xortab(t1 , t2 ) t1 = moebiustransform(t1 ) t2 = moebiustransform(t2 ) t1 + = t2 return t1 .... to facilitate the analysis and in particular to try a combinatorial study we will implement a specific presentation for equations thus obtained .
the aes algorithm takes 128 bits as input and provides 128 bits as output .
so we will have boolean functions @xmath53 .
the guiding principle is to generate a file by bit , we will have at the end 128 files .
each file containing the boolean equation of the concerned bit .
in each file , the boolean equation is presented under the form of lines containing sequences of 0 and 1 .
each line describes a monomial of the equation and the transition from one line to another means applying a ` xor ` . in order to facilitate understanding of the chosen mechanism
we describe the realization of file corresponding to one bit @xmath54 from his equation to the file formalism in figure [ fig : fichiers_bit ] page .
we will now apply to the aes the mechanism described above .
the difficulty with our approach is that the encryption functions of the aes algorithm takes 128 bits as input and provides 128 bits as output .
so we will have boolean functions @xmath53 and it is impossible to calculate their truth tables . indeed , in this case , we have @xmath55 possible combinations of 128-bit blocks and the space storage needed to archive these blocks is @xmath56 terabytes .
so we have to find a way to describe the aes encryption functions in the form of boolean functions without using their truth table .
we will now detail the solution implemented for each of the sub - functions of the aes encryption algorithm .
the function ` subbytes ` is a non - linear substitution that works on every byte of the states array using a substitution table ( s - box ) .
this function is applied independently to each byte of the input block .
so , the s - box of the aes is a function taking 8 bits as input and providing 8-bit as output .
so we can describe it as a boolean function @xmath57 .
from there , we can calculate the truth table of the s - box and use the mbius transform for obtain the normal algebraic form of the s - box . then applying the results to the 16 bytes of input block , we get 128 equations , each describing a block bit .
for example , the equation of the processing of the bit @xmath58 by the function ` subbyte ` is given in figure [ fig : bitsubbyte ] page .
in the ` shiftrows ` function , the bytes of the third column of the state table are shifted cyclically in an offset whose size is dependent on the line number .
the bytes of the first line do not suffer this offset . for this function , we do not need to calculate specific boolean function .
indeed , the only change made consists to shift bytes in the states array . in our files
, this transformation can be easily solved by using a ` xor ` .
thus , for example , the second byte of the status table becomes the sixth byte after the application of ` shiftrows ` .
this results in the following lines : 0000000000000000**10000000**00000000000000000000000000000000000000000000000000000000**01000000**00000000000000000000000000000000000000000000000000000000**00100000**00000000000000000000000000000000000000000000000000000000**00010000**00000000000000000000000000000000000000000000000000000000**00001000**00000000000000000000000000000000000000000000000000000000**00000100**00000000000000000000000000000000000000000000000000000000**00000010**00000000000000000000000000000000000000000000000000000000**00000001**0000000000000000000000000000000000000000 in the end , the equations of the function ` shiftrows ` for the 128-bit of the block @xmath59 are : @xmath60 the function ` mixcolumns ` acts on the states array , column by column , treating each column as a polynomial with four terms . each column is multiplied by a square matrix . for each column
we have : @xmath61 thus , for the first byte of the column we have the equation : @xmath62 as in @xmath63 , @xmath64 is the identity for multiplication , this equation becomes : @xmath65 we have the same simplification for all equations describing the multiplication of the column of the states array by the square matrix .
therefore we only need to calculate truth tables for multiplication by @xmath66 and @xmath67 in @xmath63 . for example
, the equations of the bits @xmath68 to @xmath58 are the following : @xmath69 to recall , in the algorithm of the aes-128 , @xmath70 words and @xmath71 words , with 1 word = 4 bytes = 32 bits .
the function ` addroundkey ` adds a round key to the state table by a simple bitwise ` xor ` operation .
these rounds keys are computed by a key expansion function .
this latter generates a set of @xmath72 words of 32 bit that to say 11 keys of 128 bits derived from the first key .
the algorithm used for the expansion of the key involves two functions ` subword ` and ` rotword ` together with a round constant ` rcon ` .
the generation of a global boolean function for the key expansion algorithm is impossible because the generation of the key for the round @xmath2 involves the key of the round @xmath73 .
this interweaving of rounds keys does not allow us to generate a global boolean function .
on the other hand it is possible to generate a boolean function corresponding to the calculation of a key of one round . the first word @xmath74 of the round key @xmath75 is calculated according to the following equation : @xmath76 with @xmath77 and @xmath78 respectively corresponding to the ` subword ` and ` rotword ` functions .
the following words @xmath79 , @xmath80 and @xmath81 are calculated according to the following equation : @xmath82 with @xmath83 .
the ` subword ` and ` rotword ` functions are built on the same principle as the ` subbytes ` and ` shiftrows ` functions , thus we can reuse the methodology finalized previously . in python language ,
the word generation function is written according to the following code ( see listing [ lst : keyword ] , p. ) . .... def
generateword(num ) : if ( num < 4 ) : w = generategenericword(wordsize*num , ' x ' ) if ( num > = 4 ) : if ( ( num % 4 ) = = 0 ) : w = generateword(3 ) w = rotword(w ) w = subword(w , rconlist[(num/4)-1 ] ) w = xorwords(w , generateword(0 ) ) else : w = generateword(num-1 ) w = xorwords(w , generateword(num%4 ) ) return w .... in this code , several scenarios are considered .
the function ` generateword ` takes in parameter the word number to generate , we know that this number is between 0 and 43
. if the number is less than 4 , the function returns the boolean identity function as the first key used by the aes is the encryption key .
if the number to modulo 4 is zero , the function returns a boolean functions describing the composition of ` subword ` and ` rotword ` functions and the application of the ` xor ` with the ` rcon ` constant .
finally , if the number to modulo 4 is not zero , the function returns the boolean function describing the ` xor ` with the corresponding word in the previous round .
we now have a boolean function describing a round expansion of the key .
as we have seen , the key expansion algorithm involves at round @xmath2 the keys of round @xmath73 . to integrate our boolean function in the encryption process of the aes , we must , at every round , add a temporary variable corresponding to the key of the previous round . as an example , the boolean equation of the bit @xmath84 of the fourth word on the 44 words generate by the key expansion process , is given in the figure [ fig : word ] page .
we have now a boolean function for each function ` subbytes ` @xmath85 , ` shiftrows ` @xmath86 and ` mixcolumns ` @xmath87 . in the arrangement of one round , these functions are combined .
so for a 128-bit block @xmath88 as output of the ` addroundkey ` function , the block @xmath89 as output of the combination of these three functions is such that : @xmath90 to realize the files as described above , it is necessary to reduce the composition of these three functions in one boolean equation . to achieve this
, we just have to replace each input variable of a function by the output value of the previous function using the following equation : @xmath91 in python language , the round generation function is written according to the following code ( see listing [ lst : round ] , p. ) .
.... def writeroundenc(numround , equasb , equasr , equamc ) : printcolor ( ' # # round%s ' % numround , green ) resultsr = [ ] resultmc = [ ] for i in xrange(blocksize ) : equasr[i ] = equasr[i].split ( ' _ ' ) resultsr.append(equasb[int(equasr[i][1 ] ) ] ) for i in xrange(blocksize ) : tmp = '' for monomial in equamc[i].split('+ ' ) : tmp + = resultsr[int(monomial.split('_')[1 ] ) ] tmp + = ' + ' resultmc.append(tmp.rstrip('+ ' ) ) binmon = generatebinarymonomes(resultmc ) return resultmc .... the boolean equation of one round of the aes for the bit @xmath84 is given in the figure [ fig : round ] page .
finally , we can now describe under the form of boolean equations the full process of aes encryption .
the function in python language computing this process is given in listing [ lst : ciphering ] page .
.... def generateencfullfiles ( ) : printcolor ( ' # # ciphering process ' , yellow ) createaesfiles('enc ' ) addroundkey(0 , ' enc ' ) writeroundenc(0 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(1 , ' enc ' ) writeroundenc(1 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(2 , ' enc ' ) writeroundenc(2 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(3 , ' enc ' ) writeroundenc(3 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(4 , ' enc ' ) writeroundenc(4 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(5 , ' enc ' ) writeroundenc(5 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(6 , ' enc ' ) writeroundenc(6 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(7 , ' enc ' ) writeroundenc(7 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(8 , ' enc ' ) writeroundenc(8 , subbytes ( ) , shiftrows ( ) , mixcolumns ( ) ) addroundkey(9 , ' enc ' ) writefinalroundenc(9 , subbytes ( ) , shiftrows ( ) ) addroundkey(10 , ' enc ' ) writeendflag('enc ' ) printcolor ( ' # # files generated ' , yellow ) .... we will now detail the solution implemented for each of the sub - functions of the aes decryption algorithm .
the aes deciphering algorithm uses the ` invshiftrows ` , ` invsubbytes ` and ` invmixcolumns ` functions .
those functions are respectively the inverse functions of ` shiftrows ` , ` subbytes ` and ` mixcolumns ` functions , used in the ciphering process .
the pseudo code of the decryption function can be written as follows ( see fig . [
fig : decipherpseudocode ] , page ) , ` nb ` corresponding to the 32-bits words numbe and ` nr ` corresponding to the rounds number used in the algorithm .
byte state[4,nb ] state @xmath92 in addrounkey(state , w[nr*nb , ( nr+1)*nb-1 ] ) invshiftrows(state ) invsubbytes(state ) addroundkey(state , w[round*nb , ( round+1)*nb-1 ] ) invmixcolumns(state )
invshiftrows(state ) invsubbytes(state ) addrounkey(state , w[0 , nb-1 ] ) state the internal mechanisms to the three functions used in the round during decryption are similar to encryption functions .
so we use the same reasoning as the one implemented earlier to generate the corresponding boolean equations .
for example , the boolean equation of the three transformations used in the deciphering process for the bit @xmath84 are given in figure [ fig : threefunctions ] page .
the key expansion function is the same for both ciphering and deciphering process .
boolean equations we built previously are reusable .
we have now a boolean equation for each of ` invsubbytes ` @xmath93 , ` invshiftrows ` @xmath94 and ` invmixcolumns ` @xmath95 functions .
however , unlike the arrangement of intermediate rounds of the encryption process , these three functions are not combined among them .
indeed , the function ` addroundkey ` no longer occurs at the end of the round but sits between ` invsubbytes ` and ` invmixcolumns ` functions .
thus , for a block @xmath88 and a key @xmath96 as input of the round , the block @xmath89 as output is such that : @xmath97 to reduce the boolean equations , we will not therefore be able to combine the equations of ` invsubbytes ` and ` invshiftrows ` . as before , to achieve this we just have to replace each input variable of a function with its output value of the previous function using the following equation : @xmath98 in python language , the round generation function is written according to the following code ( see listing [ lst : invround ] , p. ) .
.... def writerounddec(numround , equasb , equasr ) : printcolor ( ' # # round % s ' % numround , green ) resultsr = [ ] for i in xrange(blocksize ) : equasr[i ] = equasr[i].split ( ' _ ' ) resultsr.append(equasb[int(equasr[i][1 ] ) ] ) binmon = generatebinarymonomes(resultsr ) return resultsr .... as for the encryption process , we can now describe under the form of boolean equations the full process of the aes decryption .
the function in python language computing this process is given in listing [ lst : deciphering ] page .
.... def generatedecfullfiles ( ) : printcolor ( ' # # deciphering process ' , yellow ) createaesfiles('dec ' ) addroundkey(10 , ' dec ' ) writerounddec(9 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(9 , ' dec ' ) writeinvmixcolumns(9 ) writerounddec(8 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(8 , ' dec ' ) writeinvmixcolumns(8 ) writerounddec(7 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(7 , ' dec ' ) writeinvmixcolumns(7 ) writerounddec(6 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(6 , ' dec ' ) writeinvmixcolumns(6 ) writerounddec(5 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(5 , ' dec ' ) writeinvmixcolumns(5 ) writerounddec(4 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(4 , ' dec ' ) writeinvmixcolumns(4 ) writerounddec(3 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(3 , ' dec ' ) writeinvmixcolumns(3 ) writerounddec(2 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(2 , ' dec ' ) writeinvmixcolumns(2 ) writerounddec(1 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(1 , ' dec ' ) writeinvmixcolumns(1 ) writerounddec(0 , invsubbytes ( ) , invshiftrows ( ) ) addroundkey(0 , ' dec ' ) writeendflag('dec ' ) printcolor ( ' # # files generated ' , yellow ) .... we now have two systems of boolean equations corresponding to the encryption process and decryption of aes .
these two systems each have : * 128 equations , one for each bit block ; * 1280 variables for the input block ; * 1280 variables for the key .
concerning the variables of keys , the fact that we have a boolean equation by round key involve that we have a set of 128 new variables at each round that is 1280 variables for the aes-128 .
each of the variables of the @xmath2 round key being described in terms of variables of the @xmath73 round key . consequently and due to the ` xor ` bitwise operation between the round key and the bits resulting from the round function , we are obliged to insert a new set of 128 variables to describe the block transformation at each round .
finally we described the aes encryption and decryption process in the form of two systems of boolean equations with 128 equations and 2560 variables .
this mechanism allows us then to describe all of the aes encryption process in the form of files using the same representation as described above .
so we have 128 files , one by bit of block . in these files ,
each line describes a monomial and the transition from one line to the next is done by the ` xor ` operation . to implement this mechanism of the description of the aes encryption algorithm and generate the 128 files , we have developed and used a python script based on that described earlier in our presentation of aes .
the main program , ` aes_equa.py ` , offers the possibility of one hand to generate the files for aes ciphering and deciphering functions with the ` generateencfullfiles ( ) ` and ` generatedecfullfiles ( ) ` functions and on the other hand , to control that the encryption and the decryption obtained from files is consistent .
thus , the functions ` controlencfullfiles ( ) ` and ` controldecfullfiles ` performs respectively the encryption and the decryption from the previously generated files .
the function ` controlencfullfiles ( ) ` takes as input a block of 128 bits of plain text and a 128-bit block of key while the function ` controldecfullfiles ( ) ` takes as input a block of 128 bits of cipher text and a a 128-bit block of key .
the selected blocks are those provided as test vectors in appendix b of fips 197 @xcite .
the obtained results correspond to those provided in the fips : files we generated well represent the aes encryption and decryption algorithm .
the result obtained by the function ` generateencfullfiles ( ) ` is shown in figure [ lst : generate_encfiles ] page and the result obtained by the ` controlencfullfiles ( ) ` is shown in the listing [ lst : control_encfiles ] page .
the control function ` controlencfullfiles ( ) ` injects in the boolean functions the 128 initial variables corresponding to the clear text block and the 1280 variables corresponding to the key blocks of each round .
0.38 .... ./aes_equa.py # # ciphering process # # create directory aes_files # # addroundkey0 # # round0 # # addroundkey1 # # round1 # # addroundkey2 # # round2 # # addroundkey3 # # round3 # # addroundkey4 # # round4 # # addroundkey5 # # round5 # # addroundkey6 # # round6 # # addroundkey7 # # round7 # # addroundkey8 # # round8 # # addroundkey9 # # round9 # # addroundkey10 # # files generated .... 0.58 ....
./aes_equa.py # # clear block 00112233445566778899aabbccddeeff # # key block 000102030405060708090a0b0c0d0e0f # # addroundkey0 00102030405060708090a0b0c0d0e0f0 32 # # round0 5f72641557f5bc92f7be3b291db9f91a 32 # # addroundkey1 89d810e8855ace682d1843d8cb128fe4 32 # # round1 ff87968431d86a51645151fa773ad009 32 # # addroundkey2 4915598f55e5d7a0daca94fa1f0a63f7 32 # # round2 4c9c1e66f771f0762c3f868e534df256 32 # # addroundkey3 fa636a2825b339c940668a3157244d17 32 # # round3 6385b79ffc538df997be478e7547d691 32 # # addroundkey4 247240236966b3fa6ed2753288425b6c 32 # # round4 f4bcd45432e554d075f1d6c51dd03b3c 32 # # addroundkey5 c81677bc9b7ac93b25027992b0261996 32 # # round5 9816ee7400f87f556b2c049c8e5ad036 32 # # addroundkey6 c62fe109f75eedc3cc79395d84f9cf5d 32 # # round6 c57e1c159a9bd286f05f4be098c63439 32 # # addroundkey7 d1876c0f79c4300ab45594add66ff41f 32 # # round7 baa03de7a1f9b56ed5512cba5f414d23 32 # # addroundkey8 fde3bad205e5d0d73547964ef1fe37f1 32 # # round8 e9f74eec023020f61bf2ccf2353c21c7 32 # # addroundkey9 bd6e7c3df2b5779e0b61216e8b10b689 32 # # round9 7ad5fda789ef4e272bca100b3d9ff59f 32 # # addroundkey10 69c4e0d86a7b0430d8cdb78070b4c55a 32 69c4e0d86a7b0430d8cdb78070b4c55a ( fips result ) .... according to the same principle as for boolean functions of encryption , the result obtained by the function ` generatedecfullfiles ( ) ` is shown in the listing [ lst : generate_decfiles ] page and the obtained result from the ` controldecfullfiles ( ) ` function is shown in the listing [ lst : control_decfiles ] page .
0.38 .... ./aes_equa.py # # deciphering process # # create directory aes_files # # addroundkey10 # # round 9 # # addroundkey9 # # invmixcolumns 9 # # round 8 # # addroundkey8 # # invmixcolumns 8 # # round 7 # # addroundkey7 # # invmixcolumns 7 # # round 6 # # addroundkey6 # # invmixcolumns 6 # # round 5 # # addroundkey5 # # invmixcolumns 5 # # round 4 # # addroundkey4 # # invmixcolumns 4 # # round 3 # # addroundkey3 # # invmixcolumns 3 # # round 2 # # addroundkey2 # # invmixcolumns 2 # # round 1 # # addroundkey1 # # invmixcolumns 1 # # round 0 # # addroundkey0 # # files generated .... 0.58 .... ./aes_equa.py # # cipher block 69c4e0d86a7b0430d8cdb78070b4c55a # # key block 000102030405060708090a0b0c0d0e0f # # addroundkey10 7ad5fda789ef4e272bca100b3d9ff59f 32 # # round9 bd6e7c3df2b5779e0b61216e8b10b689 32 # # addroundkey9 e9f74eec023020f61bf2ccf2353c21c7 32 # # invmixcolumns9 54d990a16ba09ab596bbf40ea111702f 32 # # round8 fde3bad205e5d0d73547964ef1fe37f1 32 # # addroundkey8 baa03de7a1f9b56ed5512cba5f414d23 32 # # invmixcolumns8 3e1c22c0b6fcbf768da85067f6170495 32 # # round7 ... # # round3 fa636a2825b339c940668a3157244d17 32 # # addroundkey3 4c9c1e66f771f0762c3f868e534df256 32 # # invmixcolumns3 3bd92268fc74fb735767cbe0c0590e2d 32 # # round2 4915598f55e5d7a0daca94fa1f0a63f7 32 # # addroundkey2 ff87968431d86a51645151fa773ad009 32 # # invmixcolumns2 a7be1a6997ad739bd8c9ca451f618b61 32 # # round1 89d810e8855ace682d1843d8cb128fe4 32 # # addroundkey1 5f72641557f5bc92f7be3b291db9f91a 32 # # invmixcolumns1 6353e08c0960e104cd70b751bacad0e7 32 # # round0 00102030405060708090a0b0c0d0e0f0 32 # # addroundkey0 00112233445566778899aabbccddeeff 32 00112233445566778899aabbccddeeff ( fips result ) .... in both cases , encryption and decryption , the results we obtain by using our files to cipher and to decipher blocks are conform to those described in the fips 197 .
so our boolean equation system describing the aes algorithm is right .
after presenting briefly the boolean algebra , boolean functions and two of their presentations , we have developed a process that allows us to translate the aes encryption and decryption algorithms in boolean functions
. then we defined a mode of representation of these boolean functions in the form of computer files .
finally , we have developed a program to implement this process and to check that the expected results are consistent with those provided in the fips . in the end , we got a two new systems of boolean equations , the first one describing the entire ciphering process while the second describes the entire deciphering process of the _ advanced encryption standard _ and each one including 128 equations and @xmath99 variables
. the next step could be to search , through statistical and combinatorial analysis , new ways to cryptanalyse the aes .
either by finding a solution to resolve our equations system either by using statistical bias exploitable with this system .
michel dubois and ric filiol , _ proposal for a new equation system modelling of block ciphers and application to aes 128 _ , proceedings of the 11th european conference on information warfare and security , 2012 .
michel dubois and ric filiol , _
proposal for a new equation system modelling of block ciphers and application to aes 128 - long version _ , pioneer journal of algebra , number theory and its applications , 2012 . | one of the major issues of cryptography is the cryptanalysis of cipher algorithms .
cryptanalysis is the study of methods for obtaining the meaning of encrypted information , without access to the secret information that is normally required .
some mechanisms for breaking codes include differential cryptanalysis , advanced statistics and brute - force .
recent works also attempt to use algebraic tools to reduce the cryptanalysis of a block cipher algorithm to the resolution of a system of quadratic equations describing the ciphering structure . in our study , we will also use algebraic tools but in a new way : by using boolean functions and their properties .
a boolean function is a function from @xmath0 with @xmath1 , characterized by its truth table .
the arguments of boolean functions are binary words of length @xmath2 .
any boolean function can be represented , uniquely , by its algebraic normal form which is an equation which only contains additions modulo 2 the ` xor ` function and multiplications modulo 2 the ` and ` function .
our aim is to describe the aes algorithm as a set of boolean functions then calculate their algebraic normal forms by using the mbius transforms .
after , we use a specific representation for these equations to facilitate their analysis and particularly to try a combinatorial analysis . through this approach
we obtain a new kind of equations system .
this equations system is more easily implementable and could open new ways to cryptanalysis .
* keywords * : block cipher , boolean function , cryptanalysis , aes |
in mott insulators , the heisenberg interaction @xmath6 is in most cases the dominant source of coupling between local moments , and most theoretical investigations are based on modeling in which only this type of interaction is included .
it has been known for a very long time , however , that other , less symmetric , interactions are present . for instance , unless there is an inversion center on a bond , spin - orbit coupling induces an antisymmetric interaction of the form @xmath7 , which is known as the dzyaloshinskii - moriya ( dm ) interaction.@xcite since it breaks the fundamental su(2 ) symmetry of the heisenberg interactions , the dm interaction is at the origin of many deviations from pure heisenberg behavior , such as canting@xcite or small gaps.@xcite it is also known to have a dramatic impact on the properties of antiferromagnets in a magnetic field .
numerous experimental investigations of quantum antiferromagnets currently in progress in large field facilities call for a detailed understanding of this problem.@xcite several issues have recently been the subject of rather intensive research . for instance , the impact on triplon bose - einstein condensation@xcite of dm interactions has been analyzed.@xcite the interplay of frustration and dm interactions has also received significant attention.@xcite the consequence of the breaking of su(2 ) symmetry on the excitation spectrum is also well understood thanks to the work of several people including some of the present authors.@xcite it is by now well established that a dm interaction can open a gap in otherwise gapless regions .
the scaling of this gap with the magnitude of the dm interaction has been worked out for several cases.@xcite surprisingly , however , the other important consequence of the breaking of the su(2 ) symmetry on the ground state properties of weakly coupled dimers , namely the development of a local magnetization , has not received much attention so far , although it is of immediate relevance to several compounds .
it was shown in the case of srcu@xmath8(bo@xmath9)@xmath8 that a dm interaction can lead to the development of a measurable ( and in fact quite large ) staggered magnetization,@xcite but a simple picture of how the magnitude and the orientation of the dm interaction with respect to the magnetic field influences these properties has not yet emerged .
besides , the fact that a dm interaction can lead to the development of a transverse uniform magnetization and its impact on torque measurements of the magnetization have not been investigated in detail .
all these questions are central to the understanding of several systems of current interest .
in particular , recent nmr results by clmancey _ et al._@xcite have revealed the presence of a staggered magnetization in the dimer compound cu@xmath2(c@xmath3h@xmath4n@xmath2)@xmath2cl@xmath5 [ abbreviation : cu(hp)cl ] , and the interpretation of these results requires a precise investigation of the effect of dm interactions on weakly coupled dimer systems . in this paper , we present a systematic analysis of the consequences of an intra - dimer dm interaction on the development of local magnetization in systems of weakly coupled spin 1/2 dimers .
we first look at the case of an isolated dimer , and derive simple expressions in the limits of weak and strong magnetic field which we believe are very useful to get a simple picture of subtle effects such as the effect of the relative orientation of the magnetic field and the @xmath10 vector of the dm interaction on the uniform magnetization .
we then turn to the case of coupled dimers and concentrate on a simple ladder geometry .
this choice is motivated partly by the potential relevance of this geometry to actual compounds such as cu(hp)cl , and by the possibility to obtain very accurate results using the density matrix renormalization group method ( dmrg)@xcite in this quasi - one dimensional geometry .
finally , we report new torque measurements on cu(hp)cl and discuss them in the light of these results .
the problem of an isolated dimer in a magnetic field in the presence of a dm interaction is defined by the hamiltonian @xmath11 the @xmath12 axis has been chosen to be that of the magnetic field , and the @xmath13 plane as the plane defined by the magnetic field and the @xmath0 vector ( see fig . [
fig : model - dimer ] ) . in actual systems ,
the direction of the @xmath0 vector is fixed by the microscopic arrangement of atoms and orbitals , and it is the orientation of the magnetic field that can be varied with respect to the crystal , but the convention of having the magnetic field along the @xmath12 axis makes the discussion somewhat simpler . the @xmath0 vector is written as @xmath14 . + the isolated dimer problem is , of course , very simple .
the hilbert space is of dimension 4 , and it will prove convenient to work in the basis @xmath15 unfortunately , the explicit expressions that can be derived for the eigenenergies and eigenvectors are cumbersome and not particularly useful .
therefore , we will discuss the results from the point of view of symmetry , will derive useful expressions for small @xmath0 in weak field and close to saturation , and will present plots of some representative exact results . in a magnetic field , the su(2 )
symmetry of the heisenberg model is reduced to a u(1 ) symmetry corresponding to a rotation around the field direction . as soon as a dm interaction with a @xmath0 vector not parallel to the magnetic field is introduced ,
the rotational symmetry in spin space is completely lost .
the only symmetry that remains is the mirror symmetry with respect to the @xmath13 plane ( the plane containing the magnetic field and the @xmath0 vector ) , which exchanges sites 1 and 2 and simultaneously changes the sign of the @xmath16-component of the spin operators ( the component perpendicular to the mirror plane ) . as a consequence ,
the expectation values of local spin operators in any eigenstate of the hamiltonian satisfy the relations : @xmath17 these relations imply that the staggered magnetization per site , defined as @xmath18 , is perpendicular to the plane defined by the magnetic field and the @xmath19 vector , while the uniform magnetization per site defined by @xmath20 must lie in that plane .
if the @xmath0 vector is parallel to the field , the u(1 ) rotational symmetry is still present , which can be easily checked since @xmath21 commutes with @xmath22 .
the states @xmath23 and @xmath24 are still eigenstates with energies @xmath25 , but @xmath26 and @xmath27 get coupled .
the staggered magnetization is identically zero , while the uniform magnetization jumps abruptly from 0 to @xmath28 ( @xmath29 is the direction of the applied magnetic field ) at a critical field @xmath30 larger than its @xmath31 value @xmath32 . in the limit @xmath33 and below the saturation field @xmath34 , the ground - state wave function ,
up to second order in @xmath35 , reads : @xmath36 in the low - field limit , first - order perturbation theory in @xmath37 can be used to derive simple expressions for the expectation value of the various spin operators : @xmath38 these expressions lead to compact and suggestive expressions for the uniform and staggered magnetizations : @xmath39 as required by symmetry , the staggered magnetization is perpendicular to both the field and the @xmath0 vector .
as far as the uniform magnetization is concerned , symmetry only requires that it lies in the plane of the magnetic field and of the @xmath0 vector , but in the low field limit , eq .
( [ eq - m - hsmall ] ) shows that it is perpendicular to the @xmath0 vector .
so the uniform magnetization is in general _ not _ parallel to the magnetic field , as it would be in a system with su(2 ) symmetry , and it can in fact deviate strongly : in the limit where the @xmath0 vector becomes parallel to the field , the uniform magnetization becomes perpendicular to the magnetic field , a rather anomalous behavior that should have important consequences for torque measurements of the magnetization .
another remarkable feature of these results is that the staggered magnetization is first order in @xmath0 , while the uniform magnetization is second order .
thus , at low field the response is dominated by the staggered magnetization , as already observed in srcu(bo@xmath9)@xmath8 .
finally , let us emphasize that , as implied by eq .
[ eq - m - hsmall ] , the uniform and staggered magnetizations have universal expressions in terms of the magnetic field and of the @xmath0 vector , which are valid regardless of their orientation with respect to the lattice . at the critical field @xmath34
, one has to turn to degenerate perturbation theory since , for @xmath31 , @xmath26 and @xmath40 are degenerate . when the @xmath0 vector is not parallel to the field , these states get coupled by an off - diagonal term @xmath41 .
the ground state wave function is then simply given by @xmath42 , and the staggered magnetization per site is equal to @xmath43 .
interestingly enough , this maximal value is independent of the angle @xmath44 and does not vanish but remains quite large in the limit where @xmath45 goes to zero .
note , however , that the staggered magnetization only takes significant values close to @xmath46 , in an interval of width of the order of @xmath47 which shrinks to zero in the limit where @xmath45 goes to zero , consistent with a vanishing staggered magnetization when @xmath31 . when @xmath48 , the uniform magnetization at this field is equal to @xmath49 , which corresponds to half the polarization value .
when the angle between @xmath45 and @xmath37 is not @xmath50 , a small uniform component develops along @xmath51 due to the coupling of @xmath52 with @xmath53 .
this transverse ( with respect to the field ) uniform magnetization is given by @xmath54 .
in contrast to the small field result , it is now linear in @xmath45 , but remains much smaller than the staggered magnetization , which is of order one .
to get an idea of the accuracy of the expressions obtained at low field and close to the saturation field , we have plotted in fig .
[ fig - d0.04 ] the exact value of @xmath55 , @xmath56 and @xmath57 for a representative case ( @xmath58 and @xmath59 ) .
the small field expression is quantitatively accurate up to @xmath60 , and the width of the peak of the staggered magnetization and the maximal value of @xmath56 are indeed of order @xmath45 . ,
@xmath57 and @xmath55 for @xmath61 and @xmath62 .
the dashed lines are the analytical results derived in the limit @xmath33 .
note the difference of scale for positive and negative magnetizations . ]
in this section , our goal is to check to which extent the properties of a system of weakly coupled dimers resemble those of isolated dimers . in particular , the transition between zero magnetization and polarization takes place through an extended region of magnetic field of the order of the inter - dimer coupling , and we would like to know how the system behaves within and outside this region .
we will attack this problem numerically , and in order to perform simulations on large systems , we have chosen to work in a ladder geometry and to use the dmrg .
the model is defined by the hamiltonian @xmath63 as for the isolated dimer , the @xmath0 vector is assumed to lie in the @xmath13 plane , i.e. , @xmath64 .
our choice of an alternating @xmath0 vector from one rung to the other ( see fig . [ fig : model_ladder ] ) is motivated by symmetry considerations .
indeed , in a canonical ladder , the middle of each rung is an inversion center , and the dm interaction vanishes by symmetry .
a simple way to allow for the dm interaction to become finite without modifying the symmetry of the exchange couplings is to assume that some buckling is present along the ladder , as sketched in fig .
[ fig : model_zigzag ] . in that case , the only mirror plane that contains a bond is the @xmath65 plane , and a dm interaction with a @xmath0 vector parallel to @xmath51 is allowed by symmetry .
but , in this geometry , the presence of a @xmath66 axis ( see fig .
[ fig : model_zigzag ] ) implies that the @xmath0 vector alternates from one rung to the other .
the buckling realized in cu@xmath2(c@xmath3h@xmath4n@xmath2)@xmath2cl@xmath5 is slightly more subtle ( successive rungs are connected by an inversion symmetry in the middle of a plaquette ) , but this symmetry also implies alternating @xmath0 vectors .
note , however , that other ways of breaking the inversion symmetry of the rungs can lead to other arrangements of @xmath0 vectors .
-direction is allowed by symmetry . ]
another motivation to work with alternating @xmath0 vectors is to keep the perturbation caused by the inter - dimer coupling as small as possible . in that respect
, this choice is natural . indeed , as we have seen in the previous section , the presence of a @xmath0 vector on a rung induces a staggered magnetization . if the @xmath0 vectors of neighboring rungs @xmath67 and @xmath68 are equal , the moments @xmath69 and @xmath70 will also be equal , which is in conflict with antiferromagnetic inter - rung exchange interactions .
if , on the contrary , the @xmath0 vectors are opposite on neighboring rungs , the local moments will adopt configurations that are compatible with the exchange . with this choice of staggered @xmath0 vectors , the model possesses the following symmetries
: i ) a @xmath13 mirror plane going through the middle of the rungs ; ii ) an inversion center in the center of the plaquette formed by two consecutive rungs ; iii ) even translation symmetries along @xmath51 .
as long as these symmetries are not broken , the following relations between the expectation values of local spin operators on two neighboring rungs are expected to be satisfied : @xmath71 we thus define the staggered and uniform magnetizations per site as @xmath72 where @xmath73 is the total number of sites , and with the convention that the angle @xmath44 is positive for @xmath67 even . as in the isolated dimer case , the staggered magnetization @xmath74 is along the @xmath16 axis , while the uniform magnetization @xmath75 lies in the @xmath13 plane .
axis with the field .
inset : plot of @xmath57 as a function of @xmath76 slightly below @xmath77 , which confirms the scaling predicted in ref .
[ ] . ] component of the uniform magnetization as a function of the field for @xmath78 and various values of @xmath35 .
significant values appear between @xmath77 and @xmath79 , and far outside this interval as soon as @xmath35 is not too small .
upper panel : angle @xmath80 between the magnetic field and the uniform magnetization @xmath75 as a function of the field for @xmath78 and several values of @xmath35 .
note that @xmath81 goes to @xmath50 in the low - field limit , in agreement with the prediction for an isolated dimer ( eq . ) . ]
let us now turn to the discussion of the numerical results we have obtained for the model of eq .
( [ eq - ham ] ) .
we are interested in the regime @xmath82 .
for @xmath31 , the model is a simple ladder in a field , and the properties are well understood .
there is , of course , no staggered magnetization because of the u(1 ) symmetry , and the uniform magnetization is parallel to the field for the same reason .
it vanishes below a critical field @xmath77 , takes off with a square - root singularity , and reaches saturation with another square - root singularity at a second critical field @xmath79 .
the difference @xmath83 scales with @xmath84 .
since , apart from this scaling , the properties depend very little on @xmath84 , we quote results for a single value of @xmath84 , and having in mind the compound cu(hp)cl,@xcite we have chosen @xmath85 . for that ratio ,
the critical fields in the absence of dm interactions are given by @xmath86 and @xmath87 .
for the model with dm interaction , we have performed exact diagonalization ( ed)@xcite up to 20 sites ( 10 rungs ) , and dmrg calculations on ladders with up to 80 rungs .
the results evolve smoothly with the size , and we only quote dmrg results obtained for 80-rung clusters ( finite - size effects for the gap are discussed in the next section ) .
note that in those calculations , @xmath88 is not a good quantum number .
this is well known to reduce greatly the maximal size available to exact diagonalizations , but this also has an impact on the number of states we were able to keep during the dmrg runs . here , we diagonalize ( by the davidson method ) a matrix of size @xmath89 at each dmrg step . in a standard dmrg run where @xmath88 is a good quantum number , the matrix of the effective hamiltonian in the variational basis is block - diagonal , which can speed up the diagonalization by a factor of 10 or more .
the memory needed is also larger at fixed @xmath90 than for the standard dmrg . for those reasons ,
most of the calculations were done with up to @xmath91 states kept during 5 sweeps , and only up to @xmath92 sites .
the discarded weight was of the order of @xmath93 when we targeted two states to extract the gap , and of the order of @xmath94 or less when we targeted a single state to extract correlations .
we also performed a few runs with @xmath90 up to 800 in order to confirm that the numerical data were well converged . and
large values are achieved between @xmath77 and @xmath79 , and far outside this interval as soon as @xmath35 is not too small .
the value between @xmath77 and @xmath79 depends relatively weakly on @xmath35 and @xmath44 , and is of the same order as the maximal value in the case of an isolated dimer ( @xmath1 ) .
in contrast , the value outside this interval depends very strongly on the magnitude of @xmath95 . ]
the @xmath12 component of the magnetization is displayed in fig .
[ fig - magz ] for several values of @xmath45 and @xmath44 .
it is reminiscent of that for @xmath31 ; however , when @xmath96 , the magnetization develops as soon as the magnetic field is switched on , only reaching saturation asymptotically in the limit of infinite field .
the square root singularities are removed .
it was shown in ref .
[ ] that , at @xmath77 , the magnetization should depend on the magnitude of the @xmath45 vector as @xmath97 , in agreement with the present results ( see the inset of fig . [ fig - magz ] ) . when @xmath98 ( i.e. , @xmath99 ) , a uniform magnetization along the @xmath51 axis also develops , as in the isolated dimer case .
[ fig - phi - ang30 ] shows the magnetization along the @xmath51 axis and the angle @xmath80 between the uniform magnetization and the @xmath12 axis as a function of the magnetic field for @xmath78 . at low field ,
the uniform magnetization is orthogonal to the dm vector , again as for an isolated dimer .
the magnetization along @xmath51 is maximal between the two critical fields .
its value in that range is clearly much smaller than the component along the field ( @xmath80 becomes very small near @xmath77 ) , but this extra contribution to the uniform magnetization will produce a torque that should be detectable experimentally given the very high sensitivity of torque measurements .
the staggered magnetization along @xmath16 exhibits a kind of plateau in the intermediate phase between @xmath100 and @xmath101 ( fig .
[ fig - stmagx ] ) .
its magnitude inside the plateau is of the order of the maximal value of the isolated dimer ( @xmath1 ) , and it depends relatively weakly on @xmath45 .
in contrast , the extent of the tails outside this plateau region increases rapidly with @xmath45 . remarkably , the magnetization per spin along @xmath16 is larger than along @xmath12 up to @xmath77 and even slightly above .
note that the staggered magnetization depends essentially on the value of @xmath102 and is very weakly affected by the value of @xmath103 .
.5 cm for @xmath104 and several values of @xmath35 : @xmath105(black ) , @xmath106(red ) @xmath61(blue ) @xmath107(green ) and @xmath108 ( dmrg ) .
lower left panel : scaling of the excitation gap as a function of 1/n for d / j=0.02 .
lower right panel : scaling of the gap as a function of d / j for n=80 ( see text ) . ]
the effect of a @xmath109 breaking interaction on a ladder has been studied in ref . [ ] .
it strongly depends on the nature of the plateau phase . for the transition from the zero or full polarization to the gapless phase , the effective field theory
is expected to be the same as for the spin chain close to saturation , and the gaps at @xmath77 and @xmath79 should open as @xmath110 , as shown in ref .
[ ] . this prediction clearly agrees with the results for @xmath111 shown in fig . [
fig : gap ] ( lower right panel ) .
size effects are already very small for @xmath92 sites , as can be seen in fig .
[ fig : gap ] ( lower left panel ) . between the two critical fields ,
the gap is expected to remain finite .
( the closing of the gap in ref .
[ ] was caused by a breaking of the @xmath112 symmetry which does not occur here as there is no @xmath113 plateau when @xmath114 ) .
the effect of the @xmath12 component of @xmath45 is expected to be very small .
this is also confirmed by our dmrg results ( not shown ) .
if the uniform magnetization is not parallel to the field , it induces a torque @xmath115 on the system .
usually , such a torque is only present if the field is not along a high symmetry direction of the @xmath116-tensor .
this is the basis of torque measurements of the magnetization .
however , as shown above , a dm interaction can also induce a component of the magnetization perpendicular to the magnetic field , which should show up in torque experiments as an additional contribution .
interestingly enough , torque measurements on cu(hp)cl indeed reveal the presence of such a contribution .
experiments were carried in a resistive magnet and @xmath115 was measured up to 23 t at 410 mk .
the orientation of the crystal was adjusted so that @xmath117 at the highest values of @xmath37 , as shown in the inset of fig .
[ fig : torque ] , which fully cancels the contribution due to the anisotropy of the @xmath116 tensor .
this orientation indeed corresponds to @xmath118 [ 100 ] . in spite of this , a large additional contribution shows up between the critical fields , and extends well outside the intermediate region .
for comparison , the calculated component of the uniform magnetization perpendicular to the field of a ladder with @xmath119 and @xmath120 ( @xmath78 ) is depicted on the same plot , with scales adjusted to get the same value at @xmath77 .
the values of @xmath102 and @xmath121 are those used in ref . to fit the staggered magnetization , while the results depend very little on @xmath103 up to an overall scale factor .
the two curves are in good qualitative agreement , especially considering the fact that the only adjustable parameter is the overall scale factor . in order to go beyond this qualitative agreement
, it would be necessary to consider several additional effects .
first of all , inelastic neutron scattering data have challenged the description of this system as a simple ladder,@xcite and further couplings ( still note definitely identified ) should presumably be included .
in addition , there is a transition into a 3d ordered phase@xcite between @xmath77 and @xmath79 , and although the precise nature of the ordering is still unknown , it is very likely it will affect the uniform magnetization .
clearly , at the present stage , too little is known about these additional effects to be able to take them into account , and this is left for future investigation . for @xmath122 ( blue circles ) and torque divided by field ( experimental curve obtained on @xmath123 ) as a function of the field .
inset : torque measurement raw data ( black ) and torque divided by field ( red ) . ]
if spin 1/2 dimers are coupled in such a way that there is no inversion center at the middle of the bond , very significant modifications of the physics in a magnetic field have to be expected . indeed , unless it is forbidden by symmetry
, a dm interaction will always be present , and the analysis reported in this paper shows that even a tiny dm interaction can modify some aspects of the physics rather dramatically .
this is especially true for the staggered magnetization , which immediately acquires large values in the intermediate phase where the system gets polarized , and which can take on significant values outside this phase for physically relevant values of the dm interaction .
this is also true for the uniform magnetization as soon as the @xmath0 vector of the dm interaction and the field are neither parallel nor perpendicular . in that case
, a component of the uniform magnetization perpendicular to the magnetic field appears , which can induce a measurable torque on the sample .
this has been proven for an isolated dimer and for a ladder with staggered dm interactions , but these conclusions are expected to hold true for all coupled - dimer systems as long as the @xmath0 vectors are arranged in such a way that there is no competition with heisenberg exchange as far as the development of a staggered magnetization is concerned . it is our hope that these results will help understand some of the strange properties observed in coupled - dimer systems .
we acknowledge useful discussions with karlo penc and oleg tchernyshyov .
this work was supported by the grant - in - aids for scientific research on priority areas `` invention of anomalous quantum materials '' and for aoyama gakuin university 21st coe program from the ministry of education , culture , sports , science and technology of japan , by the swiss national fund , by manep , and by the sfb 382 . | we investigate the interplay of dzyaloshinskii - moriya interactions and an external field in spin 1/2 dimers . for isolated dimers and at low field , we derive simple expressions for the staggered and uniform magnetizations which show that the orientation of the uniform magnetization can deviate significantly from that of the external field .
in fact , in the limit where the @xmath0 vector of the dzyaloshinskii - moriya interaction is parallel to the external field , the uniform magnetization actually becomes _ perpendicular _ to the field . for larger fields ,
we show that the staggered magnetization of an isolated dimer has a maximum close to one - half the polarization , with a large maximal value of @xmath1 in the limit of very small dzyaloshinskii - moriya interaction .
we investigate the effect of inter - dimer coupling in the context of ladders with density matrix renormalization group ( dmrg ) calculations and show that , as long as the values of the dzyaloshinskii - moriya and of the exchange interaction are compatible with respect to the development of a staggered magnetization , the simple picture that emerges for isolated dimers is also valid for weakly coupled dimers with minor modifications .
the results are compared with torque measurements on cu@xmath2(c@xmath3h@xmath4n@xmath2)@xmath2cl@xmath5 . |
nematic elastomers ( nes ) @xcite are rubbery materials with the macroscopic symmetry properties of nematic liquid crystals @xcite .
in addition to the elastic degrees of freedom of ordinary rubber , nematic elastomers possess the internal , orientational degree of freedom of liquid crystals .
since nes are amorphous solids rather than fluids , their mechanical properties differ significantly from those of standard nematic liquid crystals .
the interplay between elastic and orientational degrees of freedom is responsible for several fascinating properties of nes .
for example , temperature change or illumination can change the orientational order and cause the elastomer to extend or contract by several hundred percent @xcite .
nematic elastomers display a soft elasticity @xcite characterized by vanishing shear stresses for a range of longitudinal strains applied perpendicular to the nematic direction .
they exhibit an anomalous elasticity in which certain bending and shear moduli are length - scale dependent and vanish or diverge at long length - scales @xcite .
in addition to unusual static properties , nes have an intriguing dynamic mechanical behavior . early rheology experiments on liquid crystalline elastomers found no nematic effects in nes .
later , however , several experiments observed a genuinely unconventional response to oscillatory shear .
it turns out that an internal relaxation of the nematic director leads to a dynamic mechanical softening of nes .
this behavior has been christened dynamic soft elasticity .
it makes nes interesting for device applications in areas such as mechanical vibration damping ( exploiting the fact that the mechanical loss is record - high over a wide range of temperatures and frequencies ) or acoustics where nes open the possibility of acoustic polarization ( using the fact that only particular soft shears are strongly attenuated ) analogous to the optical polarization in birefringent media .
the theoretical investigation of the dynamic - mechanical properties of nes was pioneered by terentjev , warner and coworkers ( tw ) .
references @xcite and present a detailed derivation of equations governing the long - wavelength low - frequency dynamics of nes and derive their associated mode structure , which includes non - hydrodynamic , rapidly decaying modes .
this derivation , which assumes a single relaxation time for the director , is based on the lagrangian approach @xcite to the dynamics of continuum systems in which non - dissipative forces are calculated as derivatives of an energy functional and dissipative forces as derivatives of a a rayleigh dissipation function .
the tw work focuses primarily on rheological response in both soft and semi - soft nes at zero wavenumber , and it neglects contributions to dynamical equations arising from the frank free energy for director distortions , which are higher order in wavenumber than those arising from network elasticity . as a result , as we shall see , the mode structure derived in refs .
@xcite and misses diffusive modes along symmetry directions in soft nes .
when contributions from the frank free energy are included and those from the elastic network are excluded , the tw approach reproduces the original leslie - ericksen equations @xcite of nematodynamics , excluding the rotational inertial term that is usually discarded @xcite or is missing entirely in alternative derivations @xcite of these equations . in this paper , we will present alternative approaches to deriving the equations governing the dynamics of nes
. first we will derive the exclusively hydrodynamical equations for those variables whose characteristic frequencies vanish with wavenumber .
these equations , like those for the hydrodynamics of smectic @xcite and columnar @xcite liquid crystals , exclude non - hydrodynamic director modes .
interestingly , the modes for soft nes predicted by these equations are identical in form to those of columnar liquid crystals @xcite with three pairs of sound modes and diffusive modes along symmetry directions where sound velocities vanish .
one pair of sound modes is predominantly longitudinal with non - vanishing velocity at all angles .
a second pair is predominantly transverse with a velocity that vanishes for wavevector either parallel or perpendicular to the uniaxial symmetry axis , and a third pair is completely transverse with a velocity that vanishes for wavevector parallel to the symmetry axis but is nonzero otherwise .
we then derive the phenomenological equations for the slow dynamics of all displacement and director variables in nes using the poisson - bracket formalism @xcite for obtaining the dynamics of coarse - grained variables , applied so successfully to the study of dynamical critical phenomena @xcite , for deriving phenomenological equations for any set of coarse - grained variables whose dynamics is slow on a time scale set by microscopic collision times . like tw , we assume a single relaxation time for the director .
we do , however , discuss how this constraint can be relaxed .
when applied to fluid nematic liquid crystals , the poisson - bracket formalism is equivalent to that used by the harvard group " in its derivation of hydrodynamics of nematics .
our dynamical equations for nes reduce to the equations of nematohydrodynamics derived by the harvard group when elastic rigidities vanish and to the purely hydrodynamics equations for nes when fast modes are removed . when contributions from the frank free energy are ignored , our equations are identical to those derived by tw .
we will not discuss random stresses or inhomogeneities in this paper , though they contribute static components in light scattering experiments @xcite that may obscure the observation of the ne modes we discuss here .
the outline of our paper is as follows . in sec .
[ harvardbasics ] we briefly review the general poisson - bracket formalism for obtaining coarse - grained dynamics .
section [ purehydrodynamics ] focuses on the pure hydrodynamics of nes .
we set up equations of motion for the momentum density and the elastic displacement .
we derive the appropriate elastic energy entering these equations by integrating out the director degrees of freedom .
then we compare the hydrodynamic equations to those of conventional uniaxial solids and columnar liquid crystals . after extracting the sound velocities of the modes we finally determine the full mode structure in the incompressible limit .
section [ low - frequencydynamics ] features our formulation that explicitly accounts for the dynamics of the director .
we set up equations of motion for the momentum density , the elastic displacement and the director .
these are then compared to the equations of motion for uniaxial solids and to the equations of motion derived by tw .
we determine the mode structure .
finally , we compare the so obtained modes to the modes obtained by tw .
a brief summary is given in sec .
[ concludingremarks ] .
stochastic dynamical equations for coarse - grained fields @xcite can be obtained by combining the poisson - bracket formalisms of classical mechanics @xcite , which guarantees the correct reactive couplings between fields with opposite signs under time reversal , and the langevin @xcite approach to stochastic dynamics , which provides a description of dissipative processes and noise forces .
let @xmath0 , @xmath1 be a set of coarse - grained fields whose statistical mechanics is described by a coarse - grained hamiltonian @xmath2 .
the dynamical equations for @xmath3 are first - order differential equations in time : @xmath4 here and in the following the einstein summation convention is understood .
the first term on the right hand side is a non - dissipative velocity , also known as the reactive term , that contains the poisson bracket @xmath5 of the coarse grained fields @xcite .
the reactive term couples @xmath6 to @xmath7 only if @xmath8 and @xmath9 have opposite signs under time reversal ( when external magnetic fields are zero ) .
the second term on the right hand side is a dissipative term .
@xmath10 are the components of the so - called dissipative tensor .
this tensor couples @xmath6 to @xmath7 only if @xmath8 and @xmath9 have the same sign under time reversal . if the noise term @xmath11 is present , eq .
( [ genharvard ] ) represents a stochastic or langevin equation . as such eq .
( [ genharvard ] ) may be used to set up a dynamic functional @xcite to study the effects of nonlinearities and fluctuations via dynamical field theory . in this paper
we are not interested in these effects .
hence , we focus on linearized hydrodynamic equations and pay no further attention to noise .
hydrodynamics describes the dynamics of those degrees of freedom whose characteristic frequencies @xmath12 vanish as wavenumber tends to zero . in other words hydrodynamics
focuses exclusively on the _ leading _ low - frequency , long - wavelength behavior .
all other degrees of freedom , even though they might be slow , are strictly speaking not hydrodynamic ones . in this section
we will derive the hydrodynamic equations of nes .
these equations apply for frequencies @xmath12 such that @xmath13 , where @xmath14 is the longest non - hydrodynamic decay time in the system .
as we will show in the next section , the characteristic time for director decay is in fact very slow with @xmath15 sec . so that the regime of applicability of hydrodynamics is quite small for current nes .
it is imaginable , however , that other systems will be found with shorter decay times .
there are two general classes of hydrodynamic variables : conserved variables and broken symmetry variables .
a single - component ne has the same set of conserved variables as a fluid : energy density @xmath16 , mass density @xmath17 , and momentum density @xmath18 .
it also has the same broken symmetry variables as a crystalline solid , namely three displacement variables , though strictly speaking , these variables in an elastomer are not associated with a macroscopic broken symmetry because in the absence of orientational order elastomers have the same macroscopic rotational and translational symmetry as a fluid . since it is rotational symmetry that distinguishes a nematic elastomer from an isotropic one , it could be argued that the nematic director should be a hydrodynamic variable , but as in smectic and columnar liquid crystals , the director degrees of freedom decay in microscopic times to their preferred configuration in the presence of strain and are thus not hydrodynamic variables .
elastomers differ from equilibrium crystals in at least two important ways : the first , alluded to above , is that they are not periodic and thus do not have mass - density wave order parameters whose phases act as broken symmetry hydrodynamic variables ; rather lagrangian displacement variables @xmath19 take their place .
the second is that an elastomer is permanently crosslinked : it is a classical rubber in which changes @xmath20 in mass density are locked to changes in volume such that @xmath21 and in which permeation in which there is translation of a mass - density - wave without mass motion is prohibited .
thus , mass density is not an independent hydrodynamic variable , and we are left with a total of @xmath22 independent variables and @xmath23 associated hydrodynamic modes .
these modes are heat diffusion and , depending on direction , either six propagating sound modes , four propagating sound modes and two diffusive displacement / velocity modes , or two longitudinal sound modes and diffusive displacement / velocity modes . in what follows
, we will consider only isothermal processes so that heat heat diffusion can be ignored . to derive the hydrodynamical equations for nes
, we first need the appropriate elastic free energy .
the elastic constant measuring the energy of strains in planes containing the anisotropy axis [ the shear modulus @xmath24 to be defined in eq .
( [ uniax ] ) ] vanishes as a result of the broken symmetry brought about by the establishment of nematic order .
thus a good starting point for this free energy is that of a uniaxial solid with this elastic constant simply set to zero .
this leaves certain soft directions in which distortions cost zero energy , and as in smectic and columnar liquid crystals @xcite , curvature - like terms that are quadratic in second - order spatial derivatives have to be added to ensure stability . to make contact with dynamical equations involving the director to be presented in sec .
[ low - frequencydynamics ] , rather than simply adding these terms , we find it useful to derive them from the free energy of a nematic elastomer expressed in terms of both the strain and the directors .
we will restrict our attention to harmonic distortions .
elastomers are permanently crosslinked pieces of rubber whose static elasticity is most easily described in lagrangian coordinates in which @xmath25 labels a mass point in the unstretched ( reference ) material and @xmath26 , where @xmath27 is the displacement variable , labels the position of the mass point @xmath25 in the stretched ( target ) material .
we will use lagrangian coordinates throughout this paper . we will , however on occasion make reference to eulerian coordinates in which @xmath28 specifies a position in space and @xmath29 the displacement variable at that position .
the elastic energy of a nematic elastomer can be divided into three parts : @xmath30 where @xmath31 is the usual elastic energy of a uniaxial solid , @xmath32 is the frank free energy of a nematic , and @xmath33 is the energy of coupling between strain and director distortions .
choosing the coordinate system so that the @xmath34-direction coincides with the uniaxial direction , we have to harmonic order @xcite @xmath35 here , @xmath36 are linearized components of the lagrange strain tensor @xmath37 : @xmath38 we will use the convention that indices from the beginning of the alphabet , @xmath39 , assume the values 1 and 2 whereas indices from the middle of the alphabet , @xmath40 , run from 1 to 3 . note that , compared to our work on the anomalous elasticity of nes @xcite the elastic constants in eq .
( [ uniax ] ) have a somewhat different definition that is geared towards taking the incompressible limit . here
we have arranged things so that the terms featuring @xmath41 and @xmath42 involve the trace of the strain tensor . in the incompressible limit that we will eventually take one
has @xmath43 so that @xmath41 and @xmath42 drop out . expanded to harmonic order in the deviation @xmath44 from the uniform equilibrium state @xmath45 , the frank energy reads @xmath46 where @xmath47 is the two - dimensional levi - civita symbol . for notational simplicity in the remainder of this paper
we will replace @xmath48 by @xmath49 with the understanding that it has only two components @xmath50 . with this notation , @xmath51
can be expressed as @xmath52 where @xmath53 and where @xmath54 with the einstein convention understood .
the coupling energy , finally , can be written in the form @xmath55 where once more terms beyond harmonic order have been neglected and where @xmath56 as explained above the director is not a genuine hydrodynamic variable and hence it should be integrated out of the elastic energy as long as we focus on the hydrodynamic limit .
we do so by minimizing @xmath2 over @xmath50 to find @xmath57 inserting this equation into @xmath2 and retaining only the dominant terms in gradients , we obtain @xmath58 where [ reno ] @xmath59 the superscript @xmath60 indicates that @xmath24 , @xmath61 and @xmath62 have been renormalized by director fluctuations . note that the elastic constants @xmath63 and @xmath64 are not the same as the frank splay and bend constants @xmath61 and @xmath62 .
this is in contrast to smectic - a and columnar liquid crystals in which the coefficients of quartic gradient terms in the effective elastic energy arising from the relaxation of director modes are identical to the frank elastic constants .
the modulus @xmath65 can have either sign , and there are no thermodynamic constraints preventing either @xmath66 or @xmath67 from being zero .
thus , either @xmath63 or @xmath64 in eq .
( [ reno ] ) could be zero . in this case , contributions to @xmath68 or @xmath69 in @xmath70 arising from network elasticity would have to be added to ensure stability .
except for the exceptional cases when @xmath71 , those contributions are generally smaller than the ones represented in eq .
( [ reno ] ) , and we will ignore them .
if @xmath72 vanishes , then the dependence of the elastic energy on @xmath73 drops out .
this is the famous soft elasticity .
if the condition for softness ( [ softnesscond ] ) is not strictly fulfilled and there is a small but finite remnant @xmath72 then small shears in the planes containing the anisotropy axis do cost a small energy .
this non - ideal behavior is known as semi - softness .
the energy ( [ h - u - el ] ) combines attributes of both smectic and columnar liquid crystals .
the vertical displacement @xmath74 is analogous to the displacement variable @xmath75 of a smectic liquid crystal .
it is soft for distortions in the @xmath76-direction , and a bending term @xmath77 is needed to stabilize it .
the in - plane displacements @xmath78 are analogous to those of columnar liquid crystal , which are soft for distortions in the @xmath34-direction , and a bending term @xmath79 is need to stabilize them .
we are now in a position to write down the full hydrodynamic equations for nematic elastomers . for simplicity
, we will restrict our attention to isothermal processes so that we can ignore temperature diffusion .
this leaves us with six hydrodynamical variables , the momentum density @xmath80 and the displacement @xmath81 .
these are independent variables at the reference point @xmath25 that satisfy the continuum generalizations of the usual relations for the momentum and displacement of a particle : @xmath82 , @xmath83 , and @xmath84 .
these relations yield a non - vanishing poisson bracket between @xmath85 and @xmath86 @xmath87 the @xmath88 and @xmath89 poisson brackets are zero . using these results , we obtain the equations of motion [ genstructsimple ] @xmath90 where @xmath91 is the velocity field , and @xmath92 is the coarse - grained kinetic energy .
@xmath93 is the viscosity tensor , which has five independent components .
we can parametrize the stress tensor so that the entropy production from viscous stresses takes on the same form as the elastic energy @xmath31 , @xmath94 where the @xmath95 s are low frequency viscosities .
@xmath96 is an effective viscosity that is , as its counterpart @xmath72 , renormalized by director fluctuations
. equation ( [ uniaxviscosity ] ) contains all contributions to the entropy production equation in the truly hydrodynamic limit we are considering .
nonhydrodynamic variables like the director have already relaxed to their local equilibrium values .
thermodynamic stability requires @xmath97 for @xmath98 , @xmath99 and @xmath100 .
a few observations about eqs .
( [ genstructsimple ] ) are in order .
first they are identical in form to the equations for a conventional elastic material ( without vacancy diffusion ) .
the distinction between such a material and a nematic elastomer appears only in the form of @xmath101 .
the absence of any dissipative term proportional to @xmath102 in eq .
( 3.13a ) reflects the tethered or crosslinked character of the elastomer . in non - crosslinked systems
, this equation would contain an additional dissipative term proportional to @xmath103 describing permeation .
second , these equations are expressed in lagrangian coordinates , and the derivatives @xmath104 and @xmath105 are time derivatives at constant value of the reference position @xmath25 . in eulerian coordinates , @xmath19 and @xmath18 become functions of points @xmath106 in space : @xmath107 and similarly for @xmath108 .
the time derivative @xmath109 is the time derivative in the reference frame moving with the local fluid velocity . to obtain a more explicit form for the equations of motion ,
we note that eq .
( [ eomsimple1a ] ) allows us to replace @xmath110 in eq .
( [ eomsimple1b ] ) with @xmath111 to produce the standard mechanical equation for a solid with dissipation , [ finalsimpleeom ] @xmath112 apart from the appearance of @xmath72 , which is zero in a soft ne , rather than @xmath24 and bending terms with quartic derivatives , these equations are identical to those of a uniaxial solid . upon switching to frequency space one can see that the static renormalization of the shear modulus @xmath24 , eq .
( [ softnesscond ] ) , has a dynamical generalization with a frequency dependent @xmath113 , whose real part vanishes when @xmath114 .
this form for @xmath115 is only valid in the hydrodynamic limit with @xmath13 . as we will discuss more fully in the next section , there are other terms in @xmath116 when this limit is not obeyed .
to assess the mode structure of the equations of motion ( [ finalsimpleeom ] ) , we begin with an analysis of propagating sound modes in the absence of dissipation . to keep arguments simple
, we will restrict ourselves here to ideally soft nes with @xmath117 . when @xmath118 , the bending terms can be neglected , and the dynamical equations and associated equations are identical to those of uniaxial solids .
we will return to the case when @xmath72 is nonzero in sec .
[ low - frequencydynamics ] .
the sound modes have frequencies @xmath119 where @xmath120 is the angle that @xmath121 makes with the @xmath34-axis .
the non - dissipative bending terms give rise to modes with @xmath122 along the symmetry direction .
these modes , however , mix with dissipative ones and become overdamped diffusive modes with @xmath123 .
thus , to obtain the true non - dissipative sound - mode structure , we can ignore the bending terms .
consequentially , the non - dissipative sound - mode structure is that of a uniaxial solid with @xmath124 , which it turns out is identical to that of columnar liquid crystals , though the input variables are slightly different ( there is no @xmath74 in a columnar liquid crystal , but there is an independent density ) .
the modes break up into displacements @xmath125 perpendicular to both @xmath121 and the @xmath34-axis and coupled displacements @xmath126 and @xmath74 in the plane that contains @xmath121 ( see fig . [ coordinates ] ) . in fourier space , the hydrodynamic equations are @xmath127 symmetry directions .
the @xmath128-direction is perpendicular to the plane containing the equilibrium director @xmath129 and the wavevector @xmath121 of a generic excitation .
the @xmath130-direction is perpendicular to @xmath121 and the @xmath128-direction.,width=188 ] thus , there is a transverse sound mode with velocity @xmath131 that vanishes when @xmath132 and reaches a maximum at @xmath133 .
the sound velocities @xmath134 and @xmath135 of the other modes are coupled and satisfy @xmath136\cos^2 \theta \sin^2 \theta ] \\
c_1 ^ 2 + c_2 ^ 2 & = \frac{1}{\rho}[(2 c_4 + c_3 ) \sin^2 \theta + ( c_1 + 2c_2 + c_3 ) \cos^2 \theta ] \end{aligned}\ ] ] these equations are identical to the equations satisfied by the sound velocities in a columnar system @xcite
. one of the sound modes is purely longitudinal in the limit @xmath137 ; its velocity is non - vanishing for all @xmath120 .
the second mode is like that of a smectic-@xmath138 @xcite with a sound velocity that vanishes at @xmath132 and @xmath139 .
the third mode is a purely transverse sound mode whose velocity vanishes only at @xmath140 .
figure [ soundvelocities ] plots the three sound velocities .
schematic polar plot ( arbitrary units ) of the sound velocities @xmath141 ( solid line ) , @xmath142 ( short dashes ) , and @xmath143 ( long dashes ) .
, width=302 ] having found the general sound mode structure in the non - dissipative limit , we turn to a full analysis of modes in the incompressible limit . to discuss this limit , it is useful to decompose @xmath144 into a longitudinal part @xmath145 along @xmath121 and components @xmath146 , perpendicular to both @xmath121 and @xmath34 , and @xmath147 perpendicular to @xmath121 in the plane containing @xmath121 and @xmath34 as shown in fig .
[ coordinates ] . as was the case in the dissipationless limit
, @xmath146 decouples from @xmath147 and @xmath145 . in the incompressible limit @xmath145
vanishes and we are left with @xmath148 \ , u_t \nonumber \\ & - i \omega \left[(\eta_1 + 2 \eta_4)\frac{q_{\perp}^2 q_z^2}{q^2}+ \frac{\eta_5^r}{2 } \ , \frac{(q_{\perp}^2- q_z^2)^2}{q^2 } \right ] \ ,
u_t \ , .\end{aligned}\ ] ] these equations produce propagating modes with respective frequencies [ propmodes ] @xmath149 as long as @xmath121 is not along a symmetry direction in which the sound velocity is zero . when @xmath150 , the @xmath128- and @xmath130-modes become diffusive with identical frequencies @xmath151 \ ,
q_z^2 \ , .\ ] ] as in conventional nematics @xmath152 and @xmath153
. the viscosity should be larger than the @xmath154 characteristic of fluids .
thus , we can expect that @xmath155 . in this limit
the above modes become slow and a fast diffusive mode with frequencies [ diffusivemodesttslowfast ] @xmath156 when @xmath157 , the @xmath128-modes remain propagating sound modes , but the @xmath130-modes become diffusive with frequencies @xmath158 \ , q_{\perp}^2 \ , , \ ] ] which in the limit @xmath159 reduce to [ noidea2 ] @xmath160 note that there is one diffusive and one propagating sound mode in the incompressible limit when @xmath161 .
the attenuation ( the imaginary part of q ) of the diffusive modes is proportional to @xmath162 and that of the sound modes is proportional to @xmath12 .
this is the explanation of the large difference in attenuation of the two polarizations of transverse wave found by tw that makes nes candidates for acoustic polarizers .
we will return to this issue in sec .
[ comptw ] .
equations ( [ slowboth ] ) and ( [ slowt ] ) show that one misses the slow diffusive modes if one neglects the frank energy .
it is a legitimate question , however , under what conditions these modes can be observed experimentally .
manifestly , they should be observable directly in the symmetry directions in which the respective sound velocities vanish whereas they should not be seen in directions which differ significantly from these symmetry directions .
between the two extremes there will be a crossover from slow diffusive to propagating behavior at certain crossover angles .
these angles can be estimated by comparing the magnitude of the terms in eqs .
( [ propmodes ] ) and ( [ slowboth ] ) or respectively ( [ slowt ] ) .
for @xmath163 we get for example from eqs .
( [ propmodes ] ) and ( [ slowboth ] ) that the crossover is expected at an angle @xmath164 such that @xmath165 the wave vectors in light scattering experiments on ne dynamics typically have a magnitude @xmath166 .
the elastic moduli @xmath167 and @xmath168 should be comparable to the shear modulus of rubber , @xmath169 .
hence , we estimate latexmath:[\ ] ] note that eq .
( [ wtpmrecast ] ) is essentially identical to the @xmath130-direction secular equation for conventional uniaxial solids .
tw considered the @xmath130-modes only for @xmath318 . for @xmath292
( [ wtpmrecast ] ) reduces to @xmath319 using table [ tab : notation ] find that eq .
( [ wtpmsimple ] ) is in full agreement with the result of tw for the qsv waves . before summarizing our findings ,
we now briefly return to the property of nes that makes them , as pointed out by tw , candidates for acoustic polarizers , viz .
the large difference in attenuation between the @xmath128- the @xmath130-modes in the symmetry direction where @xmath157 .
for the @xmath128-direction it follows immediately from eq .
( [ wtmrecast2 ] ) that the attenuation is proportional to @xmath12 if @xmath271 and @xmath157 .
equation ( [ wtpmsimple ] ) has two solutions for @xmath271 , namely @xmath320 and , with the proper identification @xmath321 , the fast diffusive mode eq .
( [ fastt ] ) .
if the frank energy is taken into account , the @xmath320 mode becomes the slow diffusive mode of eq .
( [ slowt ] ) . due to their diffusiveness ,
the attenuation of these @xmath130-modes is proportional to @xmath162 and hence much larger for low frequencies than the attenuation of the propagating @xmath128-mode .
this large difference in attenuation can be used , in principle , to split the @xmath130-modes from the @xmath128-modes .
nematic elastomers exhibit the remarkable phenomenon of soft or semi - soft elasticity in which the effective shear modulus @xmath72 for shears in planes containing the anisotropy axis respectively vanishes or is very small . in this paper , we have explored the dynamical consequences of this elasticity .
we derived dynamical equations , involving only the displacement , valid in the hydrodynamic limit in which frequencies and wavenumbers are respectively small compared to all characteristic microscopic inverse times and lengths in the system , and we determined that their mode structure is identical to that of columnar liquid crystals in the soft limit when @xmath322 .
we also used the poisson - bracket approach to derive dynamical equations , which contain non - hydrodynamic modes , for both director and displacement and verified that they reduced to those derived by terentjev , warner , and coworkers when contributions from the frank free energy can be ignored .
we analyzed the mode structure of these equations assuming a single relaxation time for the director , which we took to be longer than any other characteristic decay time such as the rouse time @xmath177 .
our equations , however , permit the introduction of frequency - dependent dissipative coefficients valid at frequencies higher than these inverse decay times .
rheological experiments at zero wavenumber have reported frequency - dependent storage and loss moduli that are in agreement with the predictions of the semi - soft theory that includes the director .
it would be interesting to map out the modes of nematic elastomers directly using light scattering .
it may , however , be difficult to access the true hydrodynamic limit because it applies in current nes only for frequencies of order 100 hz or less and because inherent sample inhomogeneities may lead to extra scattering that could mask the signals of the characteristic modes of a homogeneous system .
w. warner and e.m .
terentjev , _ liquid crystal elastomers _ ( clarendon press , oxford , 2003 ) h. finkelmann , h. j. koch , and g. rehage , makromol .
chem.rapid commun .
* 2 * , 317 ( 1981 ) .
p. g. de gennes in _ liquid crystals of one and two - dimensional order _ , p. 231
, edited by w. helfrich and g. heppke ( springer , new york , 1980 ) .
e. m. terentjev , j. phys .
cond . mat .
* 11 * , r239 ( 1999 ) .
g. de gennes and j. prost , _ the physics of liquid crystals _ ( clarendon press , oxford , 1993 )
. s. chandrasekhar , _ liquid crystals _ ( cambridge university press , cambridge , 1992 ) .
j. kpfer and h. finkelmann , macromol .
1353 ( 1994 ) . h. finkelmann and h. wermter , polym . mater .
, 319 ( 2000 ) .
h. finkelmann , e. nishikawa , g. g. pereira , and m. warner , phys .
. lett . * 87 * , 015501 ( 2001 ) .
l. golubovi and t. c. lubensky , phys
. rev.lett . * 63 * , 1082 ( 1989 ) .
h. finkelmann , i. kundler , e. m. terentjev , and m. warner , j. phys .
ii ( france ) * 7 * , 1059 ( 1997 ) .
g. c. verwey , m. warner , and e. m. terentjev , j. phys .
ii ( france ) * 6 * , 1273 - 1290 ( 1996 ) .
m. warner , j. mech .
solids * 47 * , 1355 ( 1999 ) .
o. stenull and t. c. lubensky , eurohys .
lett . * 61 * , 776 ( 2003 ) .
see also , x. xing and l. radzihovsky , eurohys.lett . * 61 * , 769 ( 2003 ) . o. stenull and t.
c. lubensky , to appear in phys .
e ( eprint : cond - mat/0307686 ) .
x. xing and l. radzihovsky , phys .
* 90 * , 168301 ( 2003 ) .
j. l. gallani , l. hilliou , p. martinoty , and p. keller , phys .
lett . * 72 * , 2109 ( 1994 ) ; j. l. gallani , l. hilliou , p. martinoty , f. doublet , and m. mauzac , j. phys .
ii ( france ) * 6 * , 443 ( 1996 )
. s. m. clarke , a. r. tajbakhsh , e. m. terentjev and m. warner , phys .
86 * , 4044 ( 2001 ) ; s. m. clarke , a. hotta , a. r. tajbakhsh , e. m. terentjev phys.rev .
e * 65 * , 021804 ( 2002 ) .
s. m. clarke , a. r. tajbakhsh , e. m. terentjev , c. remillat , g. r. tomlinson and j. r. house , j. appl . phys . *
89 * , 6538 ( 2001 ) .
p. stein , n. assfalg , h. finkelmann , and p. martinoty , eur .
j. e * 4 * , 255 ( 2001 ) .
a. hotta and e. m. terentjev , eur .
j. e * 10 * , 291 ( 2003 ) .
e. m. terentjev , i. v. kamotski , d. d. zakharov , and l. j. fradkin , phys .
e * 66 * , 052701(r ) ( 2002 ) ; see also l. j. fradkin , i. v. kamotski , e. m. terentjev and d. d. zakharov , proc .
a * 459 * , 2627 ( 2003 ) .
p. i. c. teixeira and m. warner , phys .
e * 60 * , 603 ( 1999 ) .
e. m. terentjev and m. warner , eur .
j. e * 4 * , 343 ( 2001 ) .
see for example , h. goldstein , _ classical mechanics _
( addison - wesley , reading , 1980 ) .
j. l. ericksen , arch .
* 4 * , 231 ( 1960 ) ; trans .
. rheol . * 5 * , 23 ( 1961 ) .
f. m. leslie , q. j. mech .
appl . math .
* 19 * , 357 ( 1966 ) ; arch .
28 * , 265 ( 1968 ) .
groupe detudes des cristaux liquides ( orsay ) , j. chem .
phys . * 51 * 816 ( 1969 ) .
p. c. martin , o. parodi , and p. s. pershan , phys .
rev . a * 6 * , 2401 ( 1972 ) .
k. kawasaki , in _ phase transitions and critical phenomena _ , vol .
2 , eds . c. domb amd m.s .
green ( academic , new york , 1976 ) .
see , e. g. , p. m. chaikin and t. c. lubensky , _ principles of condensed matter physics _ , ( cambridge university press , cambridge , 1995 ) .
hohenberg and b.i .
halperin , rev .
phys . * 49 * , 435 ( 1977 ) .
d. forster , t. c. lubensky , p. c. martin , j. swift , and p. s. persham , phys .
* 26 * , 1016 ( 1971 ) .
dieter forster , _ hydrodynamic fluctuations , broken symmetry , and correlation functions _ ( addison welsey , reading , mass . , 1983 ) ; d. forster , phys .
lett . * 32 * , 1161 ( 1974 ) .
see for example , j. z. xue , d. j. pine , s. t. milner , x. l. wu , and p. m. chaikin , phys .
a * 46 * 6550 ( 1992 ) .
see for example , h. riskin , _ the fokker - planck equation : methods of solutions and applications _ ( springer verlag , berlin , 1984 ) , and ref .
poisson brackets of coarse - grained variables require a proper interpretation .
perhaps most naturally , they can be regarded as coarse grained versions of the corresponding microscopic poisson brackets .
h. k. janssen , z. phys .
b : cond . mat .
* 23 * , 377 ( 1976 ) ; r. bausch , h. k. janssen , and h. wagner , _ ibid . _
* 24 * , 113 ( 1976 ) . c. dedominicis , j. physique c * 37 * , 247 ( 1976 ) ; c. dedominicis and l. peliti , phys .
b * 18 * , 353 ( 1978 ) .
h. k. janssen , in : _ dynamical critical phenomena and related topics _ , lecture notes in physics , vol .
104 , ed.c .
p. enz ( springer , heidelberg , 1979 ) ; h. k. janssen , in : _ from phase transition to chaos _
g.gyrgyi , i. kondor , l. sasvri , t. tl ( world scientific , singapore , 1992 ) .
j. schmidtke , w. stille , and g. strobl , macromolecules * 33 * 2922 ( 2000 ) .
m. schnstein , w. wille , and g. strobl , eur .
j. e * 5 * , 511 ( 2001 ) . beyond harmonic order , the left cauchy strain tensor @xmath323 rather than the usual right cauchy strain tensor @xmath324 couples to the director .
see t. c. lubensky , r. mukhopadhyay , l. radzihovsky , and x. j. xing , phys .
e * 66 * , 011702 ( 2002 ) .
see , e.g. , m. rubinstein and r. h. colby , _ polymer physics _
( oxford university press , oxford , 2003 ) .
h. stark and t. c. lubensky , phys . rev .
e * 67 * , 061709 ( 2003 ) .
p. g. de gennes , in _ liquid crystals of one- and two - dimensional order _ , edited by w. helfrich and g. heppke ( springer , berlin , 1980 ) . | we study the low - frequency , long - wavelength dynamics of soft and semi - soft nematic elastomers using two different but related dynamic theories .
our first formulation describes the pure hydrodynamic behavior of nematic elastomers in which the nematic director has relaxed to its equilibrium value in the presence of strain .
we find that the sound - modes structure for soft elastomers is identical to that of columnar liquid crystals .
our second formulation generalizes the derivation of the equations of nematohydrodynamics by the harvard group to nematic elastomers .
it treats the director explicitly and describes slow modes beyond the hydrodynamic limit . |
at first glance , pulsars and dark matter appear to have nothing in common , the former are magnificent spinning neutron stars with impeccable timing @xcite , while the latter embodies the most profound mystery at the crossroads of gravity and particle physics @xcite .
but , on closer inspection , one actually realises that they share more than meets the eye .
@xcite recognised this seemingly innocuous conflict when they noted that pulsars would be one of the biggest obstacle to proving a dark mater astrophysical signal .
an avalanche of recent results has just reinforced the ambiguity @xcite .
this would be purely anecdotal were it not for the fact that we have not identified a dark matter culprit .
the underlying reason is that pulsars and dark matter are predicted to share similar spectral signatures with sharp cutoffs , despite dramatically different astrophysical origins . around pulsars , gamma - ray photons
are emitted via curvature radiation of accelerated particles with an exponential cutoff at the maximum curvature energy around a few gev @xcite .
in contrast , a number of dark matter models predict that cosmic dark particles will annihilate into known elementary particles that will subsequently generate secondary photons .
the resulting gamma - ray spectrum should show a cutoff near the dark matter particle mass @xmath1 @xcite . this issue has come to bear on current searches for dark matter in the purlieus of the galactic centre .
the central concentration of dark matter is arguably the most promising place to search for unusual annihilation products .
as it turns out , over the past few years a number of groups have noticed the presence of excess gev emission around the galactic centre @xcite .
whilst these results are possible breakthroughs in dark matter research , the region over which the excess gev emission has been found is scientifically daunting with local sources of diffuse emission and unresolved gamma - ray emitters that can easily sequester any secondary emission associated with dark matter .
procedurally , a final confirmation of dark matter annihilation must exclude all other available astrophysical explanations . exploiting the spectral shape of the excess gev emission ,
@xcite and @xcite have concluded that known gamma - ray pulsars can not account for such a signal .
most arguments against pulsars have been built around the premise that the spectral shape of _ fermi _ pulsars ( @xmath2 ) can not account for the much harder ( @xmath3 ) spectrum of the gev excess @xcite .
these studies lead to the seemingly unavoidable conclusion that we are detecting dark matter annihilation .
the true situation is more complicated . in their favour
, the exponential cutoff in pulsars has been measured exquisitely well @xcite .
the observed peak for the excess gev emission at 14 gev is consistent with the observed cutoff energy for _ fermi _ pulsars that tend to cluster around 0.4 gev @xmath4 6 gev . the average _ fermi _ gamma - ray luminosity for msps from six months of _ fermi _ data also appears to be in the right ballpark of the excess @xcite . unlike dark matter , the existence of a galactic centre population of neutron stars has also been firmly established based on discovery of a handful of pulsars within @xmath5 of sgr a@xmath6 @xcite . thus , although there are some differences at the astrophysical level , observationally it is still very difficult to tell pulsars and dark matter apart for models with dark matter particle mass @xmath1 between 0.1 and 100 gev .
here we suggest that a population of hard ( @xmath0 ) msps could still account for the gev excess at the galactic centre .
initially , we estimate the pulsar population needed within a few degrees of the galactic centre in the context of the newly released second _ fermi _ large area telescope ( lat ) pulsar catalogue @xcite .
next , we motivate and discuss potential reasons for a concentration of hard msps in the innermost region .
we close with implications and possible ways forward .
our first task is to revise the measured pulsar luminosities and certify that it is possible to reproduce the excess gev emission at the galactic centre with the most recent list of _ fermi _ msps . using three years of data , the second _ fermi _ lat pulsar catalogue ( 2fpc ) reports a total of 117 gamma - ray pulsars of which 77 are young or middle - aged and 40 are msps @xcite .
the updated catalogue nearly triples the previous _ fermi _ pulsar list and it appears to be progressively populated by more msps with harder photon index @xmath0 .
this is illustrated in figure [ figure1 ] , where we plot the photon spectral index against the energy flux from 0.1 to 100 gev for _ fermi _ msps .
as in @xcite , we adopt a 0.1 100 gev energy flux of @xmath7 erg @xmath8 s@xmath9 for the gev excess from the galactic centre . at the distance of the galactic centre ( 8.3 kpc )
, this corresponds to a gamma - ray luminosity @xmath10 erg s@xmath9 @xmath11 , where @xmath11 represents the correction factor .
following @xcite , we can estimate the total number of msps potentially present in the galactic centre from the gamma - ray luminosity @xmath12 and the average luminosity of a typical pulsar @xmath13 using , @xmath14 in order to solve this relation , we must first understand the current lat sensitivity to pulsars . starting with the predicted sensitivity limits for a pulsar - like spectrum with @xmath15 and @xmath16 gev reported by @xcite
, we have built a sensitivity curve for the entire range of power law indices assuming a pulsar - like exponential cutoff energy spectrum , with a fixed parameter @xmath16 gev @xmath17 figure [ figure1 ] shows the lat sensitivity from 0.1 to 100 gev for @xmath18 and @xmath19 .
clearly , there is a strong latitude dependence of the sensitivity .
in addition , the sensitivity degrades by a factor of @xmath20 for harder pulsar spectra .
another source of uncertainty is the diffuse flux in the innermost region .
only 5 out of 40 msps are located at @xmath21 , including 3 of the most luminous msps in the entire sample . in short ,
we have not fully resolved the msp population in the galactic centre region . assuming a flux threshold of @xmath22 erg @xmath8 s@xmath9 for a photon index @xmath3 implies that @xmath23 are needed to reproduce the excess gev emission from the galactic centre . for a more conservative estimate
, we can take psr j16003053 as a typical representative of the hard - spectrum pulsar population at the galactic centre . with a luminosity @xmath13 = @xmath24 erg s@xmath9 and a photon index @xmath25
, psr j1600 - 3053 has the hardest pulsar spectrum of the msps in the 2fpc @xcite .
a population of comparable luminosity would raise the requirement to @xmath26 .
both estimates are still in agreement with the population of predicted msps ( @xmath27 ) at the galactic centre derived from observations at other wavelengths @xcite . for msps .
the solid line denotes the effective sensitivity curve for pulsar spectra with an exponential cutoff energy of @xmath16 gev at @xmath18 .
the dotted curve is for high - latitude pulsars at @xmath19.,width=336 ] this does not necessarily imply that all msps in the galactic centre region will be gamma - ray emitters . of the 169 known field radio msps , about 40 have been detected by _
therefore at least 20% of msps should be detectable in gamma rays .
but at this point , it seems that we have only scratched the tip of iceberg in terms of msp detections and could be missing the bulk of these systems @xcite .
in terms of luminosity , msps appear to be viable explanation for the inner gamma - ray excess .
however , we need to further motivate the presence of a population of hard - spectrum msps at the galactic centre . from a theoretical standpoint ,
spectral variations are expected in the photon index depending on the viewing geometry and the contribution from different emission regions @xcite .
in contrast , annihilating dark matter should be spectrally invariant across the sky .
our analysis thus far admits two possible msp scenarios .
the first is that _ fermi _ is detecting an msp population that is truly unique to the galactic centre .
based on egret observations , @xcite argued that most msp pulsars near the galactic centre are formed from old , slow moving neutron stars that have been recycled to msps .
because of its high stellar density ( @xmath28@xmath29 stars pc@xmath30 ) , only dense globular clusters ( @xmath31 stars pc@xmath30 ) with a long dynamical history can come close to mimicking the galactic centre neighbourhood .
steady gamma - ray emission has been significantly detected towards a growing population globular clusters @xcite .
these tend to show hard spectral indices ( @xmath32 ) and exponential cutoffs in the range 1.02.6 gev , which go in the right direction to explain the excess gev emission . given the similarities in stellar densities , we want to test whether the galactic centre is consistent with the properties of gamma - ray - emitting globular clusters .
one of the observational properties of globular clusters that might provide context is the apparent trend for higher gamma - ray luminosity with increasing [ fe / h ] for globular clusters @xcite .
figure [ figure2 ] marks the location where the galactic centre falls with respect to the fundamental plane of globular clusters derived by @xcite . from the figure
, we see that the galactic centre appears to be incompatible with the fundamental - plane relationship .
apart from the particular stellar dynamics around the supermassive black hole in the galactic centre , the difference may be due to a near solar metallicity of the galactic centre [ fe / h ] = 0.12 @xcite and the presence of compact young clusters in the central 50 pc that can reach central densities as high as @xmath29 stars pc@xmath30 @xcite .
this could be the first tentative indication that we are dealing with an endemic msp population .
as yet , the formation channels for msps remain a puzzle .
it is generally agreed that close , interacting x - ray binaries eventually end up as msps @xcite . using population synthesis models , @xcite argued that roche lobe overflow ( rlof ) systems involving the collapse of massive onemg white dwarfs should be pervasive in the galactic centre .
metal - rich stars fill their roche lobe more easily @xcite , and as a result the formation rate of msps could be enhanced compared to globular clusters .
this would explain the deviation of the galactic centre from the fundamental plane of globular clusters . as for the hard spectrum measured by @xcite
, we note that the median photon index of the x - ray sources discovered by _
chandra _ within the inner @xmath33 of the galaxy is @xmath34 @xcite . in the 2fpc
, one sees a growing trend @xmath35 for harder spectrum at lower msp spindown luminosity @xcite .
the spindown luminosity can be written @xmath36 where @xmath37 is the dipole moment , @xmath38 is the rotation frequency , and @xmath39 is the magnetic inclination angle @xcite .
one simple prescription is that the msp formation process near the galactic centre favours smaller magnetic inclination angles @xmath40 .
as argued by @xcite , this might be a natural tendency for recycled gamma - ray pulsars in general , but could be more frequent in the galactic centre . if recycled pulsars with onemg companions are prevalent in the galactic centre region @xcite , they will be much slower rotators ( lower @xmath41 ) than msps with he wd companions @xcite . .
the straight line shows the fundamental plane relationship from @xcite .
the dashed lines represent 95% confidence bands .
the star shows the galactic centre , consistent with the tendency for metal - rich environments and higher gamma - ray luminosity but clearly an outlier from the reported relationship.,width=336 ] since there appears to be an emerging population of hard spectra outside the galactic plane , the alternative scenario is that hard msps are not necessarily tied to the galactic centre , but have formed throughout the galaxy .
if true , the incipient hard ( @xmath0 ) sample might reveal some clues about their origin .
looking at the seven hard _ fermi _ pulsars individually , there is a smorgasbord of pulsar oddities .
six of these seven msps are in binary systems .
psr j16142230 hosts the most massive pulsar known to date @xcite .
psr j20510827 has one of the shortest orbital periods @xmath42 hr @xcite , while psr j2302 + 4442 has one of the longest @xmath43 days @xcite .
psr j16003053 is among the best high - precision pulsars known @xcite .
psr j21243358 is the lone isolated msp @xcite .
psr j01016422 is the only object where simple geometric emission models fail to explain the observed peaks , suggesting that the details of its msp magnetosphere are more complex than expected @xcite . because of frequent encounters and companion exchanges , the galactic centre could be more conducive to the production of pulsar oddities .
when compared with other field msps , no single connecting thread stands out in this bunch .
however , we can not rule out that these systems were formed in rare special environments .
dedicated studies of possible msp birth locations and companions could reveal additional information .
in view of our results , it is still possible to explain the gamma - ray excess in the galactic centre with a population of hard - spectra msps .
an essential test of these ideas is to search for similar excesses in other sections of the galaxy . interestingly , a possible excess coincident with the _ fermi _ bubbles
has been reported @xcite .
also @xcite indicated that current diffuse gamma - ray models under predict the data in the galactic plane .
it is possible that some of the msps discussed here might have migrated to regions adjacent to the galactic centre .
even if msps are not the culprits of the excess , inverse compton scattering of diffuse x - ray emission near the galactic centre @xcite by relativistic electrons ( lorentz factors @xmath44 5001000 ) could leave an imprint at gev energies . a deep galactic centre survey with the ability to resolve hard sources planned for
the cherenkov telescope array should help clarify this issue @xcite .
it seems clear that modelling the unresolved pulsar distribution will be critical step to assess astrophysical signatures of dark matter . beyond the galactic centre , pulsars above the galactic plane
could potentially mimic galactic dark matter subhalos @xcite .
nearby pulsars might also provide a source for the observed rising positron fraction @xcite . as a result , rather than treating excess gev emission as evidence for dark matter , it now seems obligatory to start including undetected pulsars as one of the largest contributors to this complex signal @xcite .
this has been done rather successfully with the diffuse gamma - ray emission from the interstellar medium @xcite .
an excellent first attempt to account for pulsars was advanced by @xcite .
since it is not obvious how to include a population that might be generally undetectable to _ fermi _ surveys , we must be extremely meticulous in building a composite pulsar template @xcite .
perhaps it will be found that pulsars and dark matter contribute to the gev excess . with improved techniques
, we could start disentangling the dark signal . from the data reported so far , it is tempting to conclude that we are starting to see the first signals of annihilating dark matter .
but we must take this road with caution in view of the degeneracy with pulsars .
technically , a population of msps with steep spectrum would be very difficult to probe in radio @xcite .
the dearth of photons near the lat sensitivity would also make it very difficult to conduct gamma - ray blind period searches with _ fermi _ @xcite .
none the less , a dedicated search for gamma - ray pulsations from the inner galaxy is a must @xcite .
a radio survey for additional pulsars with the next generation of sensitive receivers also appears to be a necessity . at the end of the day ,
the strongest astrophysical case for dark matter annihilation will be able to convene multiple sources with the same spectral signature across the sky .
a tie - break would be the localisation of `` crab - like '' power - law tails in _ fermi _ pulsars at energies above 20 gev with the upcoming cherenkov telescope array @xcite .
an alternative possibility is the direct detection of a spatially extended dark matter source @xcite .
a line - like gamma - ray feature would be a game point @xcite .
confirmation in at least two pillars of dark matter detection clinches the game @xcite .
however , indirect dark matter detection may prove much more subtle . observing the galactic centre more frequently as part of a renewed _ fermi _
observing strategy could start to break the stalemate @xcite .
we thank k. s. cheng for referring us to the @xcite paper .
we acknowledge dan hooper and per olof lindblad for helpful email exchanges .
we also thank the referee for useful suggestions and comments on the manuscript .
n.m . acknowledges support from the spanish taxpayers through a ramn y cajal fellowship and the consolider - ingenio 2010 programme under grant multidark csd2009 - 00064 . | evidence of excess gev emission nearly coinciding with the galactic centre has been interpreted as a possible signature of annihilating dark matter . in this paper
, we argue that it seems too early to discard pulsars as a viable explanation for the observed excess . on the heels of
the recently released second _ fermi _ lat pulsar catalogue ( 2fpc ) , it is still possible that a population of hard ( @xmath0 ) millisecond pulsars ( msps ) either endemic to the innermost region or part of a larger nascent collection of hard msps that appears to be emerging in the 2fpc could explain the gev excess near the galactic centre .
[ firstpage ] ( cosmology : ) dark matter
gamma - rays : observations
( stars : ) pulsars : general |
during the past several years , there has been considerable excitement about the possibility of the formation of disoriented chiral condensates ( d@xmath0c ) in heavy ion collisions @xcite .
the basic scenario is as follows : in an ultrarelativistic heavy ion collision some region thermalizes at a temperature above the chiral restoration temperature .
if the system cools sufficiently rapidly back through the transition temperature , the region will remain in a chiral restored phase .
however , this phase is unstable ; small fluctuations in any chiral direction ( @xmath1 ) will grow exponentially .
this can create regions where the pion field has macroscopic occupation . it should be stressed that this scenario is not derivable directly from the underlying theory of qcd and contains a number of untested dynamical assumptions , principally that the cooling is rapid .
thus the failure of the system to form a d@xmath0c can not be used to rule out that the system has reached the chiral restoration temperature . on the other hand , observation of the formation of a d@xmath0c would be clear evidence that the phase transition had been reached . unfortunately , since the scenario is not derived directly from a well - defined theory , it is difficult to know precisely what constitutes observation of a d@xmath0c .
assuming the system forms a single large domain of d@xmath0c containing a large number of pions there should be clear signatures . in the first place , one expects an excess in the number of low @xmath2 pions produced .
they would be at low @xmath2 since , by hypothesis , the region is large so the characteristic momentum is small ; the excess would be measured relative to a purely statistical thermal distribution .
such a signal has the advantage of working even if multiple regions of d@xmath0c form provided each region is large enough so the characteristic momentum is sufficiently small to provide a discernible signal over the thermal background .
such a signal is not decisive since one could imagine some other collective low energy effects which produce low @xmath2 pions .
a much stronger signal of a single large domain of d@xmath0c has been proposed . since the pions formed in a d@xmath0c
are essentially classical they form a coherent state .
the coherent state has some orientation in isospin space ( or more precisely the system is a quantum superposition of coherent states with different orientations and particular correlations to the isospin of the remainder of the system @xcite ) .
in essence all of the pions in the domain are pointing in the same isospin direction .
provided the total number of pions in the domain is large , this implies that the distribution of the ratio of neutral to total pions in the domain is given by @xcite @xmath3 where r is the ratio of the number of @xmath4 s in the d@xmath0c divided by the total number of pions and @xmath5 is the probability .
the derivation of @xmath5 is quite simple and will be discussed below .
the distribution in eq .
( [ pofr ] ) is qualitatively distinct from a purely statistical distribution in which the emission of charged and neutral pions is uncorrelated .
the distribution from uncorrelated emissions in the infinite particle number limit approaches a delta function at @xmath6 . for finite ( but large )
particle number the statistical distribution is narrowly peaked about 1/3 with a variance , @xmath7 where @xmath8 is the total number of pions . since these two distributions are so radically different one should in principle have a very clear signal if a single region of d@xmath0c where to form in heavy ion reactions and if the pions from the d@xmath0c are kinematically separated from other pions in the system .
the dramatic nature of the preceding signature is based in large measure on the assumption that a single large domain of d@xmath0c is formed . _ a priori _ this seems rather unlikely for the following reason : if a large region of the system starts in a hot chirally restored phase and then rapidly cools through the phase transition , then there will be a large region which is unstable against growth of the pion field .
presumably , this happens as a `` seed '' fluctuation in a small region which rapidly grows .
it takes a time of at least @xmath9 for information about the formation of the domain to propagate a distance @xmath10 .
however during the time this fluctuation is growing out to @xmath10 , the pion field at @xmath10 has been sitting in an unstable situation .
the characteristic time it can remain in this unstable configuration is @xmath11 , the exponential growth time .
if the information about the initial seed does not reach @xmath10 is a time comparable to @xmath11 the region near @xmath10 will likely begin its own exponential growth but in a chiral direction uncorrelated from the initial growth .
thus , one expects domains of characteristic size @xmath12 @xcite .
the effect of multiple domains on the @xmath13 distribution is fairly clear : it will tend to wash out the signal .
if a large number of domains form and the pions emerging from different domains can not be distinguished kinematically it is clear from the central limit theorem that the @xmath13 distribution will approach a normal distribution .
this normal distribution may be distinguished from the normal distribution arising from uncorrelated emission ; the case of multiple domains of d@xmath0c will have a substantially larger variance .
unfortunately , there is an important practical limitation which makes it difficult to exploit the @xmath13 distribution as a signature . even under the most optimistic of scenarios ,
the total number of pions coming from d@xmath0c s will be a small fraction of the total number of pions .
if one includes all pions produced in the reaction , the signal from the pions from the d@xmath0c will presumably be overwhelmed .
thus , it is highly desirable to use kinematic consideration to enhance the contributions coming from the d@xmath0c . in particular , it is sensible to study the @xmath13 distribution for a sample restricted to low @xmath2 pions only . in any scenario where the d@xmath0c is well defined , _
i.e. _ , the occupation number is large is likely to require a moderately large regions of d@xmath0c and the characteristic momentum spread in the d@xmath0c will be fixed by the inverse size of the region .
thus one expects d@xmath0cs to preferentially produce moderately low @xmath2 pions .
( one also should restrict the pions in the distribution to a moderately narrow rapidity window ) . as an experimental matter
, it should be relatively straightforward to cut on the momentum of the charged pions to select low @xmath2 pions in a given rapidity window .
for neutral pions , however , it is not a simple matter .
the neutral pions will decay in flight and will ultimately be detected as photons . if one is simply interested in the overall @xmath13 distribution , without cuts , and if the detected photons come predominately from @xmath14 decays then one can use the @xmath15 as a surrogate for @xmath16 .
recent experimental searches have exploited this strategy@xcite . however , in order to study the @xmath13 distribution in a limited kinematical region it is necessary to reconstruct the @xmath14 momenta from the observed photons in order to make kinematical cuts on the @xmath14 momenta .
since the number of neutral pions per event is large , the reconstruction of neutral pion momenta is likely to be a formidable task .
this raises the following interesting question : can one find a signature for the presence of regions of d@xmath0c of essentially the same quality as the @xmath13 distribution but which does not require the measurement of neutral pions ? in this article , we will show that the distribution of the number of charged pions ( in a kinematically limited region ) contains essentially the same information about d@xmath0c formation as the the @xmath13 distribution .
this should greatly aid in searches for d@xmath0c formation .
we also discuss additional information that can be inferred if @xmath14 s can be reconstructed .
in particular , we show that the distribution of the total number of pions ( in a limited kinematic region ) provides a means to distinguish d@xmath0c formation from other hypothetical mechanisms for the production of a pion coherent state .
we begin by studying an overly simple model and in subsequent sections we will generalize our results to more realistic scenarios . in this simplified situation
we assume that in every collision a single large domain of d@xmath0c is formed with a large particle number .
moreover , we will assume that the field strength and spatial distribution of this domain do not vary event by event , and that the pions produced in the d@xmath0c are kinematically completely distinguishable from all other pions in the system ( including those pions produced from `` @xmath17 s '' _
i.e. _ , fluctuations in the @xmath18 directions ) . finally , we will assume that both isospin violating effects and explicit chiral symmetry breaking are negligible . by hypothesis
, the region of d@xmath0c contains many particles and is essentially classical in nature . to simplify discussion
we will adopt the usual convention of describing the physics in terms of the degrees of freedom in a linear sigma model with o(4 ) symmetry , _ i.e. _ , @xmath17 and @xmath19 rather than directly in terms of the qcd degrees of freedom .
we wish to stress , however , that we are not relying on the detailed dynamics of any particular variant of the @xmath17 model .
being a coherent state , the d@xmath0c can be written in the following form : @xmath20 is the vector of creation operators in the chiral space .
the unit vector @xmath21 denotes the orientation of the d@xmath0c in the chiral space , @xmath22 is the distribution of the d@xmath0c in momentum space , and @xmath23 is the vacuum .
the number operators for neutral and charged pions are @xmath24 and one can easily find their expectation values , _
i.e. _ , @xmath25 .
it turns out that @xmath26 can be factorized into the following form .
@xmath27 where @xmath28 , and @xmath29 the expectation value @xmath30 measures the total number of @xmath31 s produced by the d@xmath0c if fully oriented in a pionic direction .
the geometrical factors @xmath32 and @xmath33 will be called the neutral and charged proportions respectively , and they take the following forms . @xmath34
note that @xmath30 does not depend on the orientation angles @xmath35 while @xmath36 do not depend on the dynamical variables @xmath37 and @xmath38 .
one can also evaluate the higher moments of @xmath39 , which gives @xmath40 with @xmath41 let s define the _ deviance _ @xmath42 $ ] of a distribution of variable @xmath43 such that @xmath44 ) \langle x \rangle^2 , \qquad \hbox{or } \qquad ( \delta x)^2 \equiv \langle x^2 \rangle - \langle x \rangle^2 = \delta[x ] \langle x \rangle^2.\ ] ] in this case , @xmath45 = 1/\langle n \rangle \to 0 \quad \hbox{when $ \langle n \rangle \to \infty$}.\ ] ] actually it is straightforward to show that @xmath46 is described by a poisson distribution , which always has a small deviance as the variance is proportional to the mean .
the above analysis shows that , _ for each set of orientation angles @xmath35 , _ the distributions of @xmath39 are normal .
however , since the orientation is randomly generated in the process of spontaneous symmetry breaking , one can not predict @xmath35 . on the other hand ,
since we are neglecting explicit chiral symmetry breaking , the system is equally probable to point in any direction in chiral space .
moreover , since by hypothesis we are in a semiclassical situation ( large @xmath30 ) , it is legitimate to work with probabilities rather than amplitudes . using the technology of ref .
@xcite , the probability distribution in the angular variables is given by the unit measure : @xmath47 one can use eq .
( [ defg ] ) to reparametrize the probability distribution ( [ pa ] ) in terms of the neutral and charged proportions @xmath48 , _
i.e. _ , the fraction of neutral or charged pions among all particles produced by the d@xmath0c .
@xmath49 from this one can obtain the marginal probability distribution of @xmath50 and @xmath51 by integrating over the other variable .
@xmath52 these distribution functions are plotted in fig . 1 in solid curves .
it is obvious the both distributions are far from being normal .
the function @xmath53 is heavily skewed towards the low end and actually diverges as @xmath54 when @xmath55 . on the other hand ,
@xmath56 is flat , and @xmath51 is equally likely to assume any value between 0 and @xmath57 .
this is drastically different from pion emission from an uncorrelated source ( the dotted curves in fig . 1 and @xmath58 for neutral and charged pions respectively .
the variances depend on the number of independently emitted pions ; the plots correspond to the case of @xmath59 .
] ) , where both distributions would be normal .
more quantitatively , one can calculate the first and second moments of @xmath50 and @xmath51 .
@xmath60 and hence the respective deviance @xmath61 $ ] and @xmath62 $ ] , @xmath63 = 1 , \qquad \delta[g_\pm ] = 1/3.\ ] ] what does the distributions of the proportions @xmath36 tell us about the distributions of @xmath39 ?
it is obvious that when @xmath46 is fixed , @xmath36 give the pion distribution . in reality , of course , @xmath46 is not fixed ; we have shown that it behaves like a poisson distribution if we have a conventional glauber coherent state @xcite . however , since poisson distributions are sharply peaked , the dispersion of @xmath46 will simply smear the distribution of @xmath64 slightly , without changing the overall shape qualitatively . more specifically ,
note that are statistical in nature , while that of @xmath46 is via quantum mechanical smearing of a coherent state .
consequently the distributions of @xmath46 and @xmath36 are assumed to be uncorrelated , and we derived the equalities below .
] @xmath65)\ , ( 1+\delta[g_{0,\pm } ] ) \ ; ( \langle n \rangle \ , \langle g_{0,\pm } \rangle)^2 \equiv ( 1+\delta[n_{0,\pm } ] ) \langle n_{0,\pm } \rangle^2 .
\end{aligned}\ ] ] the first equality gives @xmath66 which have the simple interpretation that , given the symmetry between the four directions @xmath67 in the chiral space , a quarter of the particles produced by the d@xmath0c will be @xmath4 , while half of them will be @xmath68 or @xmath69 . on the other hand , the second equality gives @xmath70 = \delta[n ] + \delta[g_{0,\pm } ] + \delta[n]\delta[g_{0,\pm}].\ ] ] with @xmath71 = 1/\langle n\rangle$ ] , @xmath61=1 $ ] and @xmath62 $ ] , one has @xmath72 = 1 + 2/\langle n \rangle , \qquad \delta[n_\pm ] = 1/3 + 4/3\langle n \rangle .
\label{ds}\ ] ] as expected , when @xmath30 is large , the deviances @xmath73 $ ] approach @xmath74 $ ] .
note that both @xmath73 $ ] are of order 1 , in contrast to an uncorrelated emission which would have @xmath75 . in an ideal world , such enhancements of @xmath76 s would indicate the existence of d@xmath0c .
lastly , it is also useful to reproduce the aforementioned probability distribution of @xmath77 by reparametrizing distribution ( [ j ] ) in terms of @xmath13 and @xmath78 ( the subscript @xmath79 stands for total ) . @xmath80 with the marginal distributions @xmath81 note that while @xmath53 is drastically skewed towards the low end and @xmath56 is flat , @xmath82 is skewed towards the high end .
this may sound counter - intuitive , but one must bear in mind that the emissions of neutral and charged pions are not independent and there is no contradiction . also note that @xmath83 is exactly as predicted in ref .
@xcite ( cf .
( [ pofr ] ) ) .
we have studied the simple case of pion emission from a single huge ( in the sense that @xmath30 is large ) d@xmath0c domain . as discussed in the introduction ,
this scenario is presumably not realistic .
we will now proceed to study more realistic scenarios with multi - domain formation .
the main point is , even though the probability distribution is smeared out because of the lack of alignment ( in the chiral space ) between the different domains , one feature survives , namely the large variance of the distributions . in particular
, we will see that the variances for both neutral and charged productions are still much larger than that of an uncorrelated emission .
let s consider the case which we have @xmath84 domains , each with the same @xmath30 .
we are also assuming both @xmath30 and @xmath84 are much larger than unity .
the total number of pions of each species produced is the sum of pions of that particular species produced in each domain , the distribution of which has been discussed in the previous section .
@xmath85 by the central limit theorem , the probability distribution of @xmath86 will approach normal distributions when @xmath84 is large .
however , we will see that the variances of the gaussian distributions will be much larger for pion production from a d@xmath0c than those of uncorrelated pion emission .
since the pion production in each domain are independent , @xmath87 are independent random variables .
hence the mean of @xmath88 is just the sum of the means of all @xmath87 , @xmath89 and the variance of the sum @xmath39 is just the sum of the variances of each @xmath87 .
@xmath90 = ( \delta[n_{0,\pm } ] / n ) { \cal n}_{0,\pm}^2 .
\label{var}\end{aligned}\ ] ] in other words , @xmath91 = { ( \delta \sigma n_{0,\pm})^2 \over \langle \sigma n_{0,\pm } \rangle^2 } = { \delta[n_{0,\pm } ] \over n } = { \delta[n_{0,\pm } ] \langle n_{0,\pm } \rangle \over { \cal n}_{0,\pm}}. \label{wide}\ ] ] in comparison with uncorrelated pion production , with @xmath92 we see that for multi - domain d@xmath0c the deviances are enhanced by a factor of @xmath93 \langle n_{0,\pm } \rangle$ ] .
@xmath94 _ a priori _ @xmath30 can take any value , but @xmath95 are larger than unity for any value of @xmath96 .
even for a very modest @xmath97 , @xmath98 and @xmath99 , leading to substantial widening of the corresponding distributions , an observable signature of d@xmath0c formation . for larger values of @xmath30 , the broadening will be even more pronounced . while the above scenario describes d@xmath0c with multi - domains , a probable feature of d@xmath0c formation in the real world ( if it happens at all ) , it is still unrealistic in assuming all the domains are of equal strength , _
i.e. _ , with the same @xmath30 . instead one expects @xmath30 of different domains to fall under a certain probability distribution , which depends on the details of the model
. naturally , one questions if the signatures discussed above still survive under such circumstances .
let s consider the case with @xmath84 domains , with different @xmath100 .
equation ( [ mean ] ) becomes @xmath101 where @xmath102 is the average of @xmath103 .
equation ( [ var ] ) becomes @xmath104.\ ] ] by the inequalities @xmath105 > \delta[g_{0,\pm}]$ ] ( cf .
eq.([ds ] ) ) and @xmath106 ( mean of squares is larger than square of mean ) , we have @xmath107 \geq \left ( \sum_{i=1}^n \langle n^{(i)}_{0,\pm}\rangle \right)^2 \delta[g_{0,\pm } ] / n = ( \delta[g_{0,\pm}]/n ) { \cal n}_{0,\pm}^2.\ ] ] in other words , @xmath91 = { ( \delta \sigma n_{0,\pm})^2 \over \langle \sigma n_{0,\pm}\rangle^2 } \geq { \delta[g_{0,\pm } ] \over n } = { \delta[g_{0,\pm } ] \overline { \langle n_{0,\pm } \rangle } \over { \cal n}_{0,\pm}}.\ ] ] ( compare eq .
( [ wide ] ) . )
again , the deviances are much larger than that of uncorrelated emission with @xmath108 when @xmath109 by the following enhancement factors .
@xmath110 so we see that , even with domains of unequal strengths , the number of neutral or charged pions produced by a d@xmath0c will still have a much wider distribution than that from independent , uncorrelated emission .
lastly , one may also ask if the distribution of @xmath88 will approach normal distributions when @xmath84 is large in the case of domains with unequal strengths . in this case
@xmath111 s do not all fall under the same probability distribution and the most simple form of the central limit theorem does not apply . on the other hand , there are generalized forms of the central limit theorem , which state that as long as the probability distributions are sufficiently `` well behaved '' , the sum of @xmath84 random variables will still fall under a normal distribution when @xmath112 .
it is actually possible to argue that the distribution of @xmath113 does approach a normal distribution by the lindeberg generalization of the central limit theorem .
( see , for example , sec .
6.e of ref . @xcite . ) whether the same conclusion holds for @xmath114 is still an open question .
let us recapitulate what we have shown in the previous sections .
we have shown that one can calculate the probability distribution of the number of neutral or charged pions produced as a result of d@xmath0c formation .
the resultant deviances @xmath76 s are not of the order @xmath115 as in an uncorrelated emission , but are instead enhanced by factors @xmath116 s which are of order @xmath30 . seeing such enhancements of deviances would be signatures of coherent pion productions .
one can understand the origin of such enhancements of statistical fluctuations of the number of neutral or charged pions by considering the following analogy .
consider two groups of gamblers playing roulette in a casino : @xmath84 lawyers at $ 100 tables , and @xmath117 physicists at the $ 1 tables , where the odds are the same . if each lawyer and physicist is given @xmath46 chips , of $ 100 and $ 1 , respectively ( so that the total amount given to the lawyers , @xmath118 , is the same as that given to the physicists , @xmath119 ) , and is required to bet all of them , the average loss will be the same for both groups as long as they are following the same betting strategies .
however , it is easy to see that the statistical fluctuation of the loss of the lawyers would be much larger for that of the physicists . in other words ,
the amount of loss , as well as its standard deviation , is `` quantized '' in units of the value of the bets .
the larger the bet , the larger the fluctuation . on the other hand , one can also turn the argument around ; a discerning external observer can deduce , with the knowledge of the gambling strategies , the size of the bets of the lawyers from the statistical fluctuations of the lawyers losses , and do likewise for the physicists as well . just as the chips in a single bet of a lawyer share the same fate ( either win or lose ) , all the pions in a single d@xmath0c domain share the same orientation in the chiral space . as a result ,
the fluctuation of the number of pions in each direction in the chiral space is enhanced by a factor proportonal to @xmath30 , the number of pions in each d@xmath0c domain . and by reverse argument
, one can deduce whether coherent pion emission is taking place by measuring the fluctuation of the number of emitted pions .
our analysis has assumed that all the pions originate from coherent emissions , and each of these coherent states has @xmath120 . in the real world
, there is contamination from independent pion emissions , and the total numbers of neutral or charged pions are the sums of these two contributions .
@xmath121 where `` c '' and `` inc '' stands for `` coherent '' and `` incoherent '' , respectively .
the variances of the numbers of coherently produced neutral or charged pions are enhanced while those of incoherent production are not .
@xmath122 where @xmath95 are of order @xmath123 ( cf . eq . ( [ e1 ] ) , ( [ e2 ] ) )
. then one can calculate the variance of the sum of the two contributions .
@xmath124 with @xmath125 the parameter @xmath0 measures the fraction of pions which are coherently produced : @xmath126 when all the pions are from d@xmath0c , while @xmath127 when all of them are independently emitted .
obviously , the more incoherent pions in the sample , the smaller is the enhancement factor @xmath128 .
while these incoherently emitted pions dilute our signatures for d@xmath0c , they have different momentum spectra from those from d@xmath0c .
d@xmath0c pions , being produced from coherent state , carry low @xmath2 .
the typical @xmath2 is of the order of @xmath129 , where @xmath10 is the size of the domain from which the pion originates .
in contrast incoherently emitted pions can carry high @xmath2 . therefore applying a low @xmath2 cut
can minimize the noise from incoherent pion emissions .
it is also advantageous to measure the rapidity of the pions and count their numbers in narrow rapidity windows .
bear in mind that the rapidities of the pions are , up to small dispersions , equal to that of the original domain . as we have mentioned , it is probable that many domains are formed in a single collision , and all these domains may have different rapidities .
for example , the domains at the surface of the `` fire ball '' are moving with high speed relative to the domains at the center . by binning the pions according to their rapidities , one can partially separate the pions from different domains , and the signals are enhanced as a result .
the above procedure allows us to search for signatures from d@xmath0cs by counting only the charged pions .
this is important as , with our present technology , it is difficult to count the number of @xmath4 s in a momentum bin , which would mean reconstructing all the pions from photons a formidable task . on the other hand , with great experimental effort
, it may be possible to count the neutral pions as well in the future . in that case
, we will be able to distinguish d@xmath0c formation from other mechanisms of coherent pion productions .
for example , one can count @xmath133 , the number of pions ( both neutral and charged ) in each rapidity window . for d@xmath0c formation , or any other mechanisms of coherent pion productions where
the field is aligned with a random direction in the four - dimensional chiral space @xmath134 , the fluctuation of @xmath133 is large . on the other hand , for mechanisms of coherent pion productions where the field is aligned with a random direction in the three - dimensional isospace @xmath135 but without involving the @xmath17 direction ,
it is straightforward to show that the fluctuation of @xmath133 is small . in summary
, we have constructed signatures for d@xmath0c formation in heavy ion collisions which do not require counting the number of neutral pions .
instead , we suggest counting the number of charged pions produced , and a large fluctuation would be a signal of d@xmath0c formation .
we believe these new signatures will be useful in searches for d@xmath0c at rhic and lhc .
andreev , jetp lett . * 33 * 367 ( 1981 ) . v. karmanov and a. kudrjavtsev , itep-88 ( 1983 ) .
anselm , phys .
* b217 * 169 ( 1988 ) .
anselm and m.g .
ryskin , phys .
* b266 * 482 ( 1991 ) .
blaizot and a. krzywicki , phys . rev .
* d46 * 246 ( 1992 ) .
bjorken , int .
j. mod .
* a7 * 4189 ( 1992 ) .
bjorken , acta phys .
polon . * b23 * 561 ( 1992 )
. k. rajagopal and f. wilczek , nucl .
* b399 * 395 ( 1993 ) .
kowalski and c.c .
taylor , hep - ph/9211282 .
k. rajagopal and f. wilczek , nucl .
* b404 * 577 ( 1993 ) .
bjorken , k.l . kowalski and c.c .
taylor , proceedings of les rencontres de physique de la valle daoste , la thuile ( 1993 ) .
bjorken , k.l . kowalski and c.c .
taylor , proceedings of the workshop on physics at current accelerator and the supercollider , argonne , il ( 1993 ) s. gavin , a. gocksch and r.d .
pisarski , phys .
lett . * 72 * 2143 ( 1994 ) .
s. gavin and b. muller , phys . lett .
* b329 * 486 ( 1994 ) .
cohen , m.k .
banerjee , m. nielsen , x.m .
jin , phys .
* b333 * 166 ( 1994 ) .
f. cooper , y. kluger , e. mottola and j.p .
paz , phys .
* d51 : * 2377 ( 1995 ) .
m. asakawa , z. huang , x. n. wang , phys . rev .
lett . * 74 * 3126 ( 1995 ) .
k. rajagopal , proceeding on quark
gluon plasma 2 , euegne , or ( 1995 ) .
cohen , phys .
* b372 * 193 ( 1996 ) .
bjorken , acta phys .
* b28 * 2773 ( 1997 ) .
minimax collaboration , poceedings of 1996 annual divisional meeting ( dpf 96 ) of the division of particles and fields of the american physical society , minneapolis , mn ( 1996 ) .
bjorken , hep - ph/9610379 .
wa98 collaboration , phys .
* b420 * 169 ( 1998 ) .
glauber , phys .
rev . * 130 * 2529 ( 1963 ) .
fraser , `` probabaility and statistics : theory and applications '' , duxbury press ( 1976 ) . | we show that the variance in the number of charged pions ( in a suitable range of momentum space ) provides a signature for the observation of a disoriented chiral condensate ( d@xmath0c )
. the signal should be observable even if multiple domains of d@xmath0c form provided the average number of pions per domain is significantly large than unity .
the variance of the number charged pions alone provides a signal which can be used even if the number of neutral pions can not be measured in a given detector .
if the neutrals can be measured , however , the fluctuations in the total number of pions provides a signature which distinguishes disoriented chiral condensates from other hypothetical sources of coherent states of pions . |
one of the most remarkable geometric features of a closed oriented surface is a natural symplectic structure on the moduli spaces of representations of its fundamental group , as shown by goldman @xcite .
these spaces consist of conjugacy classes of homomorphisms from the fundamental group , @xmath11 , to appropriate lie groups @xmath12 .
the symplectic structure on @xmath13 incorporates the classical khler forms on the teichmller spaces ( @xmath14 ) , on the jacobi varieties of riemann surfaces ( @xmath15 ) , and on the narasimhan
seshadri moduli spaces of semistable vector bundles of rank @xmath16 ( where @xmath17 ) .
this also incorporates the atiyah
bott symplectic structure on @xmath13 determined by a compact lie group @xmath12 and a non - degenerate @xmath18-invariant symmetric bilinear form on the lie algebra of @xmath12 . for a compact oriented surface with boundary ,
one obtains a weaker structure , namely , a poisson bracket on @xmath13 , or , more precisely , a poisson bracket in the algebra of conjugation - invariant smooth functions on the representation space @xmath19 , see @xcite .
the latter bracket extends to a natural quasi - poisson bracket in the algebra of all smooth functions on @xmath19 , see @xcite .
similar results hold for the general linear group @xmath20 over any commutative ring provided @xmath19 is treated as an affine algebraic set and smooth functions are traded for regular functions , see @xcite .
this approach also works for non - compact surfaces . in the present paper , we extend these constructions to manifolds of arbitrary dimensions .
namely , we construct a bracket in representation algebras associated with the pontryagin algebra of any smooth oriented manifold with non - empty boundary . for surfaces ,
our bracket is the quasi - poisson bracket on the representation space @xmath21 mentioned above . in higher dimensions ,
the representation algebras are graded , and our bracket is a gerstenhaber bracket ( i.e. , it satisfies the axioms of a poisson bracket with appropriate signs ) . in the rest of the introduction we focus on manifolds of dimension at least @xmath22 .
to proceed , we formulate the concept of a representation algebra following @xcite .
fix an integer @xmath16 and a field @xmath23 which will be the ground field of the algebras .
given an algebra @xmath7 and a commutative algebra @xmath24 , consider the set @xmath25 of all algebra homomorphisms from @xmath7 to the algebra @xmath26 of @xmath27-matrices over @xmath24 .
each @xmath28 and each pair of indices @xmath29 determine a @xmath24-valued function @xmath30 on @xmath25 which evaluates a homomorphism @xmath31 at @xmath32 and takes the @xmath33-th entry of the resulting matrix .
these functions are the coordinates `` on @xmath25 , generating an algebra of polynomial '' @xmath24-valued functions on @xmath25 .
these coordinates satisfy various polynomial relations some of which are universal , i.e. , hold for all @xmath24 . by definition , the @xmath34-th _ representation algebra _
@xmath35 of @xmath7 is generated by the symbols @xmath36 subject to those universal relations .
one of the universal relations says that the generators commute , so that @xmath35 is a commutative algebra . for every commutative algebra @xmath24 , the algebra @xmath35 projects onto the algebra of polynomial @xmath24-valued functions on @xmath25 .
we view @xmath35 as a collective representative of these polynomial algebras . if @xmath7 is graded , then @xmath35 acquires a natural grading .
van den bergh @xcite introduced a systematic method producing poisson brackets in the representation algebras @xmath35 for all @xmath16 .
it is based on a notion of a _
double poisson bracket _ in @xmath7 defined as a linear map @xmath37 satisfying certain axioms parallel to the usual axioms of a poisson bracket in an algebra .
van den bergh shows that a double poisson bracket in @xmath7 induces a poisson bracket in @xmath35 for all @xmath16 .
we will use the term _ bibracket _ for the version of double brackets used in this paper .
further , van den bergh s construction extends to graded algebras : a _ gerstenhaber bibracket _ @xmath6 in a graded algebra @xmath7 induces a gerstenhaber bracket @xmath38 in the representation algebra @xmath35 for all @xmath16 . in terms of the generators of @xmath35 , this bracket is computed as follows : for any @xmath39 , @xmath40 , and any finite expansion @xmath41 , we have + @xmath42 the bracket @xmath38 in @xmath35 is invariant under natural actions of the group @xmath43 and the lie algebra @xmath44 on @xmath35 for all @xmath16 .
consider now a smooth connected oriented manifold @xmath0 of dimension @xmath45 with @xmath46 and with base point @xmath47 .
let @xmath48 be the space of loops in @xmath0 based at @xmath49 .
the graded vector space @xmath50 carries an associative multiplication induced by concatenation of loops ( the pontryagin product ) .
this makes @xmath7 into a graded algebra called the _
pontryagin algebra _ of @xmath0 .
we define a canonical bibracket in @xmath7 called the _ intersection bibracket_. it is constructed as follows .
pick an embedded path @xmath51\hookrightarrow \partial m $ ] connecting the base point @xmath52 to another point @xmath53 .
consider arbitrary singular cycles @xmath54 and @xmath55 .
let @xmath56 be the set of all tuples @xmath57 such @xmath58 .
each such tuple @xmath59 determines two loops in @xmath0 based at @xmath60 .
the first loop goes along @xmath61 from @xmath62 to @xmath63 , then along the path @xmath64 from @xmath65 to @xmath66 and finally along @xmath67 to @xmath68 .
the second loop goes along the path @xmath67 from @xmath69 to @xmath70 , then along @xmath64 to @xmath71 and finally along @xmath72 to @xmath52 . under appropriate transversality assumptions on @xmath73 and @xmath74 ,
the resulting map @xmath75 is a singular cycle of dimension @xmath76 . passing to homology classes and using the isomorphism @xmath77 determined by @xmath61
, we obtain the intersection bibracket in the algebra @xmath7 .
our main result is the following . +
* theorem . * _ the intersection bibracket is a well - defined gerstenhaber bibracket in the pontryagin algebra . it is natural with respect to diffeomorphisms of manifolds preserving the orientation and the base point .
_ + by the above mentioned general theory , the intersection bibracket in @xmath7 induces a natural gerstenhaber bracket in @xmath35 for all @xmath16 . in the case of a simply connected manifold @xmath0 and a field @xmath78 of characteristic zero , the milnor
moore theorem identifies the pontryagin algebra @xmath7 with the universal enveloping algebra of the graded lie algebra @xmath79 ( with the degree shifted by 1 and the whitehead bracket in the role of the lie bracket ) . in this case , the algebras @xmath35 are nothing but the representation algebras of @xmath80 .
the intersection bibracket generalizes the bibracket in the group algebra of the fundamental group of a surface introduced in @xcite .
the situation here is parallel to the string topology of chas and sullivan @xcite which generalizes to higher dimensions the goldman lie bracket for free loops on a surface . despite the simplicity of the underlying idea ,
a precise definition of the intersection bibracket requires considerable efforts and occupies about two thirds of the paper .
first of all , we introduce a version of singular homology in which the role of simplices is played by manifolds with corners .
homology theories based on manifolds with corners were implicit already in @xcite and were since considered by other authors , see , for example , @xcite and @xcite .
however , we are not aware of a systematic exposition sufficient for our aims and consequently develop our own approach .
for any topological space @xmath81 , we define _ polychains _ in @xmath81 as oriented manifolds with corners endowed with an identification of some faces and with a map to @xmath81 compatible with this identification .
the connected components of a polychain are equipped with @xmath82-valued weights which replace the coefficients of singular simplices in singular chains .
we define a _ reduction _ of polychains which eliminates redundant connected components ( like , for example , components of weight zero ) .
each polychain in @xmath81 has a well - defined reduced boundary .
if it is void , then the polychain is said to be a _
polycycle_. the polycycles in @xmath81 considered up to disjoint unions with reduced boundaries form a graded vector space @xmath83 called the _
face homology _ of @xmath81 .
the usual singular homology @xmath84 is a split quotient of @xmath83 .
next , we define smooth polychains in the loop space @xmath85 of @xmath0 and study transversality for them .
we show that all face homology classes of @xmath86 can be represented by smooth polycycles , and all pairs of face homology classes of @xmath87 can be represented by transversal smooth polycycles .
this allows us to carry out the intersection construction above and obtain a linear map @xmath88 this map descends to a linear map in singular homology @xmath89 where @xmath90 . finally , the k " unneth theorem allows us to rewrite the latter as a linear map @xmath91 which turns out to be a gerstenhaber bibracket .
the condition that @xmath82 is a field is needed only in the k " unneth theorem ; most of our exposition is therefore given over an arbitrary commutative ring . replacing loops based at @xmath60 with paths in @xmath0 connecting any points of @xmath92 , we obtain a notion of a _ path homology category _ of @xmath0 .
the intersection bibracket in the pontryagin algebra and the induced gerstenhaber brackets in representation algebras naturally extend to this wider framework .
given a smooth connected oriented manifold @xmath93 without boundary , we can remove the interior of a smooth closed ball in @xmath93 and apply our constructions to the resulting manifold with spherical boundary .
the ( isomorphism class of the ) intersection bibracket in its pontryagin algebra and the induced gerstenhaber brackets are invariants of @xmath93 . under additional assumptions on @xmath93
, we also obtain a so - called @xmath94-poisson structure @xcite on the pontryagin algebra of @xmath93 itself .
this work suggests a number of questions .
is the intersection bibracket sensitive to the smooth structure ? can it be generalized to pl - manifolds or to topological manifolds ? can it be defined in the spirit of @xcite in homotopy - theoretic terms ?
note that the technique of face homology allows to define all the chas sullivan operations @xcite .
it would be useful to formally identify the resulting geometric operations with those in @xcite .
also , it would be interesting to relate the intersection bibracket introduced in this paper to the algebraic constructions of @xcite .
the paper is organized as follows .
sections [ sect - prelim][extenscatmain ] are purely algebraic .
we recall the basics on representation algebras and brackets ( section [ sect - prelim ] ) , formulate the notions of bibrackets and gerstenhaber bibrackets ( section [ bibrackets ] ) , show how to adjust these notions in the setting of unital algebras ( section [ remremrem000 ] ) and how to extend them to categories ( section [ extenscatmain ] ) . in sections [ polychains and homology ] and
[ sect - rell ] we define face homology of topological spaces and relate it to singular homology .
sections [ smooth111 ] and [ sect - transs ] are devoted to polychains in path spaces ; this includes a study of smoothness and transversality of polychains . in section [ operation_d ]
we construct the intersection polychain .
the induced operations in face homology and singular homology are discussed in sections [ upsilon - tilde ] and [ upsilon ] respectively .
the intersection bibracket is constructed in section [ bilh ] and the key jacobi - type identity for it is verified in section [ jacobi ] .
we give sample computations of the intersection bibracket in section [ sec - examples ] and compute the induced scalar intersection form in section [ scalar_intersection_form ] . in section [ reduction ] , we discuss properties of the intersection bibracket related to the hopf algebra structure of the pontryagin algebra .
section [ kill_boundary ] treats manifolds without boundary .
the paper ends with four appendices where we discuss further algebraic properties of bibrackets , the hamiltonian reduction for bibrackets , and relate the intersection bibracket to the chas
sullivan string bracket in loop homology .
+ _ conventions .
_ the letter @xmath95 will denote a fixed commutative ring and will replace the ground field @xmath78 of the introduction . by a module ( respectively , an algebra , a linear map ) we mean a @xmath95-module ( respectively , a @xmath95-algebra , a @xmath95-linear map ) . by singular homology of a space we mean singular homology with coefficients in @xmath95 .
given a smooth oriented manifold @xmath0 and a smooth orientable submanifold @xmath96 , an orientation of the normal bundle of @xmath34 in @xmath0 determines an orientation of @xmath34 , and vice versa , via the following rule : a positive frame in the normal bundle of @xmath34 followed by a positive frame in the tangent bundle of @xmath34 is a positive frame in the tangent bundle of @xmath0 .
if @xmath97 , then the orientation of @xmath0 induces an orientation of @xmath98 using the outward vector first " rule .
+ _ acknowledgements .
_ part of this work was done while g. massuyeau was visiting the department of mathematics at bloomington in spring 2013 ; he would like to thank indiana university for hospitality and support . the work of v. turaev was partially supported by the nsf grant dms-1202335 .
the authors thank f. eshmatov for an explanation of the work @xcite .
we start by recalling relevant notions in algebra and fix our terminology . by a we mean a @xmath99-graded module @xmath100 .
an element @xmath32 of @xmath7 is if @xmath101 for some @xmath102 ; we write then @xmath103 and call @xmath104 the of @xmath32 . by definition ,
the degree of @xmath105 is an arbitrary integer . for any @xmath106 , the _ @xmath107-degree _ @xmath108 of a homogeneous element @xmath28 is @xmath109 .
a is a graded module @xmath110 endowed with an associative bilinear multiplication such that @xmath111 for all @xmath112 .
note that if the product of @xmath113 homogeneous elements @xmath114 of @xmath115 is non - zero , then the degree of this product is equal to @xmath116 . if @xmath117 , then we set @xmath118 . similarly , for @xmath106 , we write @xmath119 for @xmath120 .
we do not require a graded algebra @xmath7 to have a unit element .
if @xmath121 for some homogeneous @xmath39 , then one says that @xmath32 and @xmath122 _
commute_. for a graded algebra @xmath7 , we denote by @xmath123 $ ] the graded submodule of @xmath7 spanned by the vectors @xmath124 where @xmath125 run over all homogeneous elements of @xmath115 .
the graded algebra @xmath7 is if @xmath123=0 $ ] .
factoring any graded algebra @xmath7 by the 2-sided ideal generated by @xmath123 $ ] we obtain a commutative graded algebra @xmath126 .
given graded algebras @xmath7 and @xmath24 , by a @xmath127 we mean a degree - preserving algebra homomorphism from @xmath7 to @xmath128 .
we will consider any @xmath129-graded module @xmath130 as a @xmath99-graded module by setting @xmath131 for all @xmath132 .
any graded algebra @xmath7 determines an infinite sequence of graded algebras @xmath133 as follows , cf@xmath134 @xcite .
the graded algebra @xmath135 with @xmath16 is defined by the generators @xmath30 , where @xmath32 runs over all elements of @xmath7 and @xmath136 run over @xmath137 , and the following relations : for all @xmath39 , @xmath138 , and @xmath139 , @xmath140 in the latter formula and in the sequel we always sum up over repeating indices and drop the summation sign . a typical element of @xmath135 is represented by a non - commutative polynomial in the generators with zero free term .
the grading in @xmath141 is defined by @xmath142 for all @xmath101 .
the construction of @xmath141 is functorial : a graded algebra homomorphism @xmath143 induces a graded algebra homomorphism @xmath144 by @xmath145 for all @xmath28 , @xmath146 .
for @xmath147 we have @xmath148 and @xmath149 .
the importance of @xmath135 is due to the following fact . for any graded algebra @xmath24 ,
let @xmath150 be the graded algebra of @xmath151-matrices with entries in @xmath24 .
( a matrix has a grading @xmath152 whenever all its entries belong to @xmath153 . )
then there is a canonical bijection @xmath154 which is natural in @xmath7 and @xmath24 . here
@xmath155 stands for the category of graded algebras and graded algebra homomorphisms .
the bijection carries a graded algebra homomorphism @xmath156 to the map @xmath157 sending any @xmath158 to the @xmath151-matrix @xmath159 .
the inverse bijection carries a graded algebra homomorphism @xmath160 to the graded algebra homomorphism @xmath161 sending a generator @xmath30 to the @xmath33-th term of the matrix @xmath162 for all @xmath28 .
consequently , the endofunctor @xmath163 of @xmath164 is left adjoint to the endofunctor @xmath165 of @xmath164 .
the commutative graded algebra @xmath166 is obtained from @xmath141 by adding the relations @xmath167 for any homogeneous @xmath39 and any @xmath168 .
we call @xmath35 the _ @xmath34-th representation algebra _ of @xmath7 .
the construction of @xmath169 is functorial : a morphism @xmath143 in @xmath164 induces a morphism @xmath144 in @xmath164 , which in its turn induces a morphism @xmath170 in the category of commutative graded algebras @xmath171 . for any commutative graded algebra @xmath24 , @xmath172 consequently
, the functor @xmath173 is left adjoint to the functor @xmath174 .
let @xmath7 be a graded module and @xmath175 . by a in @xmath7 we simply mean a linear map @xmath176 .
a bracket @xmath38 in @xmath7 if @xmath177 for all @xmath112 .
a bracket @xmath38 in @xmath7 is if for all homogeneous @xmath178 , @xmath179 a bracket @xmath38 in @xmath7 satisfies the if @xmath180 for all homogeneous @xmath181 .
a degree @xmath107 bracket @xmath38 in @xmath115 satisfying and is called a , and the pair @xmath182 is called then a . for example , any graded algebra @xmath7 gives rise to a @xmath183-graded lie algebra of derivations in @xmath7 .
recall that a _ derivation in @xmath7 of degree _
@xmath184 is a linear map @xmath185 such that @xmath186 for any @xmath152 and @xmath187 for any homogeneous @xmath158 and any @xmath188 .
derivations of @xmath7 of degree @xmath189 form a module @xmath190 .
the graded module @xmath191 carries a @xmath183-graded lie bracket defined by @xmath192= \delta_1 \delta_2-(-1)^{k_1k_2 } \delta_2
\delta_1 $ ] for any derivations @xmath193 and @xmath194 of @xmath7 of degrees @xmath195 and @xmath196 respectively .
a bracket @xmath197 in a graded algebra @xmath7 satisfies the if for all homogeneous @xmath198 , @xmath199 a in a graded algebra @xmath7 is a @xmath107-graded lie bracket @xmath200 in @xmath7 which satisfies the @xmath107-graded leibniz rules .
the pair @xmath201 is called then a .
for example , any graded algebra @xmath7 is a gerstenhaber algebra of degree @xmath183 with respect to the @xmath183-graded lie bracket ( called the ) defined by @xmath202 for homogeneous @xmath39 and extended to all @xmath39 by linearity .
in this section we extend van den bergh s theory @xcite of double brackets in algebras to the setting of graded algebras .
such an extension is outlined in ( * ? ? ?
* section 2.7 ) for @xmath203 and makes uses of koszul s sign conventions . throughout this section
we fix a graded algebra @xmath7 and an integer @xmath107 .
we can expand any @xmath204 as a sum @xmath205 where @xmath206 are homogeneous elements of @xmath7 and the index @xmath207 runs over a finite set . to simplify notation , we will drop the summation sign and the index and write simply @xmath208 . similarly ,
an element @xmath209 of @xmath210 will be written as @xmath211 with homogeneous @xmath212 . unless explicitly stated otherwise , we endow @xmath213 with the `` outer '' @xmath7-bimodule structure defined by @xmath214 for any @xmath178 and @xmath215 .
we shall also use the `` inner '' @xmath7-bimodule structure on @xmath213 defined by @xmath216 for homogeneous @xmath178 and any @xmath215 . given a permutation @xmath217 of @xmath218 with @xmath219 , we denote by @xmath220 the _ graded permutation _
@xmath221 carrying any @xmath222 with homogeneous @xmath223 to @xmath224 where @xmath225 is the sum of the products @xmath226 over all pairs of indices @xmath227 such that @xmath228 .
for any @xmath106 , we similarly define the @xmath229 using the @xmath107-degree @xmath230 instead of the usual degree @xmath231 . a _ bibracket _ in @xmath7 is a linear map @xmath232 .
a bibracket @xmath6 in @xmath7 has if for any integers @xmath233 , @xmath234 a _ @xmath107-graded bibracket _ in @xmath7 is a bibracket @xmath6 in @xmath7 of degree @xmath107
satisfying the following : for all homogeneous @xmath198 , @xmath235 the following lemma shows that a @xmath107-graded bibracket in @xmath7 induces certain brackets in the representation algebras @xmath236 .
[ fdbtopb ] given a @xmath107-graded bibracket @xmath6 in @xmath7 and an integer @xmath16 , there is a unique bracket @xmath200 in @xmath237 satisfying the @xmath107-graded leibniz rules , and such that @xmath238 for all @xmath39 and @xmath239 .
the bracket @xmath200 has degree @xmath107 .
we extend to a bilinear form @xmath240 satisfying and . to see that this form is well - defined , we need to verify the compatibility with the defining relations of @xmath35 . that the right - hand side of is linear in @xmath32 and @xmath122 follows from the linearity of @xmath241 .
we now verify the compatibility with the third relation in .
pick any homogeneous @xmath242 and set @xmath243 and @xmath244
. then @xmath245 therefore , for any @xmath246 , @xmath247 to check that @xmath248 , set @xmath249 and @xmath250 .
then @xmath251 therefore @xmath252 the last claim of the lemma follows from the definitions .
consider the linear involutions @xmath253 and @xmath254 of @xmath213 determined by the permutation @xmath255 as in section [ conventions ] : for homogeneous @xmath39 , we have @xmath256 and @xmath257 . given @xmath258 , the of @xmath259 is @xmath260 .
[ bibrtranspose ] a bibracket @xmath6 in @xmath7 satisfies the first @xmath107-graded leibniz rule if and only if its @xmath107-transpose @xmath261 satisfies the second @xmath107-graded leibniz rule .
assume that @xmath6 satifies .
pick homogeneous @xmath262 and set @xmath263 , @xmath264 .
then @xmath265 so , @xmath261 satifies .
the converse is shown by a similar computation .
a bibracket @xmath6 in @xmath7 is if @xmath266 . by lemma [ bibrtranspose ] , a @xmath107-antisymmetric bibracket satisfies if and only if it satisfies .
note for the record , that given a @xmath107-antisymmetric bibracket @xmath6 in @xmath7 , we have for any homogeneous @xmath39 , @xmath267 [ fdbtopbsymm ] if in lemma [ fdbtopb ] the bibracket @xmath6 is @xmath107-antisymmetric , then the bracket @xmath268 in @xmath35 is @xmath107-antisymmetric , i.e. , satisfies .
pick any homogeneous @xmath39 and set @xmath269 .
then @xmath270 the bracket in @xmath35 constructed in lemma [ fdbtopb ] may not satisfy the @xmath107-graded jacobi identity . to compute the deviation from this identity
, we observe that any bibracket @xmath6 in @xmath7 induces a linear endomorphism @xmath271 of @xmath272 , called the , by @xmath273 where @xmath274 are as defined in section [ conventions ] .
[ qp ] let @xmath275 .
if @xmath6 is a @xmath107-antisymmetric @xmath107-graded bibracket in @xmath7 , then the associated bracket @xmath268 in @xmath35 satisfies @xmath276 for any homogeneous @xmath198 , any @xmath277 .
it follows from the definitions that @xmath278 using the commutativity of @xmath35 , we deduce that @xmath279 applying the transpositions @xmath280 , @xmath281 , and @xmath282 , we obtain @xmath283 equalities and allow us to expand the right - hand side of the formula claimed in the lemma .
we next expand the left - hand side of this formula .
set @xmath284 and observe that @xmath285 we rewrite the second summand as follows . since @xmath6 has degree @xmath107 ,
@xmath286 the commutativity of @xmath35 implies that @xmath287 the @xmath107-antisymmetry of @xmath6 allows us to compute @xmath288 from @xmath289 : by , we have @xmath290 . hence , @xmath291 as a result
, we obtain that @xmath292 cyclically permuting @xmath293 and the indices , we obtain @xmath294 and @xmath295 the required formula directly follows from the equalities . a in the graded algebra @xmath7 is a @xmath107-antisymmetric @xmath107-graded bibracket @xmath6 in @xmath7 which induces a zero tribracket .
the pair @xmath296 is called then a _
double gerstenhaber algebra of degree @xmath107_. this structure was introduced by van den bergh ( * ? ? ?
* section 2.7 ) in the case @xmath203 ; see also @xcite in the setting of differential graded algebras .
[ importantgers ] for any gerstenhaber bibracket of degree @xmath107 in @xmath7 and any integer @xmath16 , the bracket @xmath38 in @xmath35 given by lemma [ fdbtopb ] makes @xmath35 into a gerstenhaber algebra of degree @xmath107 .
this follows from lemmas [ fdbtopb ] , [ fdbtopbsymm ] , and [ qp ] . in particular , the equality @xmath297 provided by lemma [ qp ] implies the @xmath107-graded jacobi identity in which @xmath293 are replaced with @xmath298 , respectively .
we discuss modifications of the representation algebras relevant in the study of unital algebras and specifically in the study of universal enveloping algebras .
a graded algebra @xmath7 is _ unital _ if it has a two - sided unit @xmath299 .
unital graded algebras and graded algebra homomorphisms carrying @xmath300 to @xmath300 form a category @xmath301 . for a unital graded algebra @xmath115 , we define a sequence of unital graded algebras @xmath302 for @xmath16 , @xmath303 is obtained from the algebra @xmath141 defined in section [ amt0 ] as follows .
first , we adjoin a unit to @xmath141 , that is consider the unital graded algebra @xmath304 with two - sided unit @xmath305 . by definition , @xmath303 is the quotient of @xmath306 by the relations @xmath307 where @xmath308 is the kronecker delta and @xmath136 run over @xmath309 . for any @xmath310
, the bijection induces a natural bijection @xmath311 similarly , let @xmath312 be the category of commutative unital graded algebras and graded algebra homomorphisms carrying @xmath300 to @xmath300 .
set @xmath313 .
then for any @xmath314 , we have a natural bijection @xmath315 we call @xmath316 the _ @xmath34-th unital representation algebra _ of @xmath7 . from the viewpoint of algebraic geometry , this exhibits @xmath317 as the `` coordinate algebra '' of the `` affine scheme '' whose set of @xmath24-points is the set of algebra homomorphisms @xmath31 for any @xmath314 . here
`` affine schemes '' are understood in the graded category , i.e@xmath134 as representable functors from @xmath312 to the category of sets .
the same graded algebra @xmath316 can be obtained from @xmath35 by adjoining a two - sided unit @xmath305 and quotienting by the relations @xmath307 where @xmath136 run over @xmath309 . for @xmath147
, we have @xmath318 and @xmath319 . [ importantgerspp ]
let @xmath6 be a @xmath107-graded bibracket in a unital graded algebra @xmath7 and let @xmath38 be the induced bracket in @xmath35 as in lemma [ fdbtopb ] .
then there is a unique bracket @xmath320 in @xmath316 such that the projection @xmath321 is bracket - preserving .
if @xmath6 is a gerstenhaber bibracket of degree @xmath107 , then @xmath320 makes @xmath317 into a unital gerstenhaber algebra of degree @xmath107 .
denote the projection @xmath321 by @xmath102 .
clearly , @xmath102 is onto which implies the uniqueness of @xmath320 . to prove the existence , we extend @xmath38 to a bracket @xmath322 in the algebra @xmath323 by @xmath324 .
the latter bracket is easily checked to satisfy the @xmath107-graded leibniz rules , .
therefore , it suffices to verify that @xmath325 annhilates @xmath322 both on the left and on the right for all @xmath136 .
the leibniz rule for @xmath6 implies that @xmath326 .
therefore for any @xmath188 and @xmath327 , @xmath328 since @xmath329 is generated by the set @xmath330 , the @xmath107-graded leibniz rules , imply that @xmath331 .
similarly , @xmath332 .
the last claim of the lemma follows from lemma [ importantgers ] .
the @xmath333 of a @xmath183-graded lie algebra @xmath334 is the quotient of the graded tensor algebra @xmath335 by the 2-sided ideal generated by the vectors @xmath336 where @xmath125 run over all homogeneous elements of @xmath337 .
the graded tensor algebra is unital and so is @xmath333 . for any unital graded algebra @xmath338 ,
the composition with the natural linear map @xmath339 determines a bijection @xmath340 where @xmath341 denotes the category of @xmath183-graded lie algebras and @xmath338 , equipped with the commutator bracket , is regarded as an object in the category @xmath341 .
section [ remremrem ] yields for each @xmath16 , a commutative unital graded algebra @xmath342 called the _ @xmath34-th representation algebra of @xmath337_. by and , for any @xmath314 , we have a natural bijection @xmath343 though we shall not use it , note that @xmath344 is generated by the commuting symbols @xmath30 where @xmath32 runs over homogeneous elements of @xmath345 and @xmath346 run over @xmath347 , subject to the first two of the relations and the relation @xmath348 for all homogeneous @xmath349 and all @xmath346 .
we extend the definitions and results of the previous sections from graded algebras to graded categories .
we follow van den bergh ( * ? ? ?
* section 7 ) who defined a version of @xmath316 for non - graded categories with finite sets of objects . in analogy with non - unital algebras , we might consider categories without identity morphisms " .
however , we do not use such generalized categories and consider only categories in the usual sense of the word . by a _ graded category _
, we mean a small category @xmath350 such that for any objects @xmath351 of @xmath350 , the set @xmath352 is a graded module , the identity morphisms of all objects are homogeneous of degree zero , and the composition of morphisms is bilinear and degree - additive .
the latter condition means that for any homogeneous @xmath353 and @xmath354 , the morphism @xmath355 is homogeneous of degree @xmath356 . with a graded category
@xmath350 we associate a graded algebra @xmath357 where @xmath358 is the direct sum of graded modules .
the product @xmath359 of @xmath353 and @xmath360 is equal to @xmath361 if @xmath362 and to zero otherwise . for @xmath363 , the identity morphism of
@xmath81 represents an element of @xmath110 denoted @xmath364
. clearly , @xmath365 and @xmath366 for @xmath367 .
if the set @xmath368 is finite , then @xmath369 is a two - sided unit of @xmath7 ; if the set @xmath368 is infinite , then @xmath110 is not unital . for each integer @xmath16
, we introduce a unital graded algebra @xmath370 .
consider the unital graded algebra @xmath304 obtained by adjoining the two - sided unit @xmath305 to the graded algebra @xmath141 associated with @xmath371 in section [ amt0 ] .
let @xmath370 be the quotient of @xmath304 by the 2-sided ideal generated by the set @xmath372 where @xmath81 runs over all objects of @xmath350 and @xmath373 .
the algebra @xmath370 has the following universality property . for each unital graded algebra @xmath24 ,
consider the graded algebra @xmath150 of @xmath151-matrices over @xmath24 as a graded category with a single object ( a matrix has the degree @xmath102 if all entries belong to @xmath374 ) .
there is a natural bijection @xmath375 where @xmath301 is the category of unital graded algebras and @xmath376 is the set of degree - preserving linear functors @xmath377 .
note that such functors can be interpreted as @xmath34-dimensional @xmath24-representations of the category @xmath350 .
the graded algebra @xmath378 plays a similar role in the category @xmath312 of commutative unital graded algebras : for any @xmath379 , there is a natural bijection @xmath380 the same algebra @xmath381 can be obtained from @xmath35 in the same way as @xmath382 from @xmath141 .
let @xmath107 be an integer .
@xmath107-graded bibracket _ in a graded category @xmath350 is a @xmath107-graded bibracket @xmath6 in the graded algebra @xmath383 such that @xmath384 for all @xmath363 .
if such a bibracket in @xmath110 is a gerstenhaber bibracket of degree @xmath107 , then the pair @xmath385 is called a _
double gerstenhaber category of degree @xmath107_. [ bracketcategories ] let @xmath6 be a @xmath107-graded bibracket in a graded category @xmath350 .
then * for any @xmath386 , @xmath387 * for any integer @xmath16 , the bracket in @xmath35 determined by lemma [ fdbtopb ] induces a bracket @xmath38 in @xmath388 satisfying the @xmath107-graded leibniz rules , ; * if @xmath385 is a double gerstenhaber category of degree @xmath107 , then the pair @xmath389 is a unital gerstenhaber algebra of degree @xmath107 for all @xmath16 .
using the leibniz rules and the identity @xmath384 , we obtain that for any @xmath390 , @xmath391 , @xmath392 other claims of the lemma follow from the definitions and lemma [ importantgers ] .
we conclude that a double gerstenhaber category @xmath385 of degree @xmath107 gives rise to a system of gerstenhaber algebras @xmath393 of degree @xmath107 .
moreover , for any full subcategory @xmath394 of @xmath395 , the algebra @xmath396 may be viewed as a subalgebra of @xmath397 in the obvious way .
the first claim of lemma [ bracketcategories ] implies that the bibracket @xmath6 in @xmath7 restricts to a bibracket in @xmath398 . in this way
, @xmath399 becomes a double gerstenhaber category of degree @xmath107 .
in particular , any object @xmath81 of @xmath395 determines a full subcategory @xmath400 of @xmath395 consisting of @xmath81 and all its endomorphisms .
then @xmath401 is the unital graded algebra @xmath402 , the restriction of @xmath6 to @xmath403 is a gerstenhaber bibracket of degree @xmath107 , and @xmath404 .
we introduce so - called `` face homology '' of a topological space formulated in terms of maps from manifolds with faces to the space .
face homology will be instrumental in our construction of a bibracket in loop homology .
we begin by recalling the basics of the theory of manifolds with faces .
we start with a bigger class of , see @xcite , @xcite , @xcite , @xcite , and @xcite .
@xmath405-dimensional manifold with corners _ with @xmath406 , or ,
shorter , an _
@xmath405-manifold with corners _ , is a paracompact hausdorff topological space locally differentiably ( @xmath407 ) modelled on open subsets of @xmath408 . for a definition in terms of local coordinate systems and for further details ,
see @xcite . the underlying topological space of an @xmath405-manifold with corners @xmath409 is an @xmath405-dimensional topological manifold with boundary . the topological boundary of @xmath409 is denoted by @xmath410 ( the symbol @xmath410 has a different meaning in @xcite ) .
the @xmath411 carries a point of @xmath409 represented by a tuple @xmath412 in a local coordinate system to the number of non - zero terms in this tuple ( this number does not depend on the choice of the local coordinate system ) . for @xmath413 , the set @xmath414 is a closed subset of @xmath409 .
it is clear that @xmath415 also , @xmath416 is a discrete set , and @xmath417 is a smooth @xmath418-dimensional manifold for all @xmath419 .
the set @xmath420 is an open subset of @xmath410 and any @xmath421 belongs to the closure of at most @xmath422 connected components of @xmath423 .
we call @xmath409 a if @xmath409 is compact and every @xmath421 belongs to the closure of precisely @xmath422 different components of @xmath423 .
this condition implies that the closure in @xmath409 of any component of @xmath423 is an @xmath424-dimensional manifold with faces whose dimension function is the restriction of @xmath425 .
we call the closure of a component of @xmath423 a of @xmath409
. we can now define recursively on @xmath426 the notion of a face of @xmath409 . by definition , a _ face _ of @xmath409
is a connected component of @xmath409 , or a principal face of @xmath409 , or a face of a principal face of @xmath409 .
clearly , @xmath409 has only a finite number of faces , and each face of @xmath409 is a connected manifold with faces .
the union of faces of @xmath409 of dimension @xmath427 is equal to @xmath428 for all @xmath413 .
the faces of @xmath409 contained in @xmath410 are said to be .
every point @xmath209 of @xmath409 lies in the interior of a unique face @xmath429 of @xmath409 .
if @xmath430 , then @xmath429 is the closure of the component of @xmath417 containing @xmath209 for @xmath431 . if @xmath432 , then @xmath433 .
note that @xmath429 is the smallest face of @xmath409 containing @xmath209 : any face of @xmath409 containing @xmath209 contains @xmath429 as a face .
for example , any compact smooth manifold @xmath0 is a manifold with faces , and its faces are the components of @xmath0 and of @xmath92 . for any @xmath406 ,
an @xmath405-dimensional simplex is a manifold with faces and its faces are the usual combinatorial faces .
finite disjoint unions and finite products of manifolds with faces are manifolds with faces in the obvious way .
an empty set is considered as an @xmath405-manifold with faces for any @xmath406 .
following @xcite , we call a map @xmath434 from an @xmath405-manifold with faces @xmath409 to an @xmath435-manifold with faces @xmath337 _ smooth _ if , restricting @xmath259 to any local coordinate systems in these manifolds , we obtain a map that extends to a @xmath407-map from an open subset of @xmath436 to @xmath437 .
( such a map @xmath259 is said to be `` weakly smooth '' in @xcite . )
a smooth map @xmath438 is continuous and its restriction to any face @xmath439 of @xmath409 is a smooth map @xmath440 .
a map @xmath438 is a _ diffeomorphism _ if it is a bijection and both @xmath259 and @xmath441 are smooth .
diffeomorphisms of manifolds with faces preserve the dimension function and carry faces onto faces .
we can define smooth ( @xmath407 ) triangulations of a manifold with faces repeating word for word the standard definition of a smooth triangulation of an ordinary manifold ( * ? ? ?
* section 8.3 ) and requiring all faces to be subcomplexes .
( the latter condition is probably satisfied automatically but we prefer to spell it out . )
the standard methods of the theory of smooth triangulations ( * ? ? ?
* section 10.6 ) apply in this setting and show that all manifolds with faces have smooth triangulations .
a manifold with faces @xmath409 is if its underlying topological manifold is oriented . the oriented manifold with faces obtained from @xmath409 by inverting the orientation is denoted by @xmath442 . by a @xmath443 on a manifold with faces
@xmath409 we mean a partition of the set of faces of @xmath409 into disjoint subsets , called , and a family of diffeomorphisms @xmath444 numerated by ordered pairs @xmath445 of faces of @xmath409 of the same type such that * @xmath446 for any face @xmath439 of @xmath409 and @xmath447 for any faces @xmath448 of @xmath409 of the same type ; * if @xmath449 are faces of @xmath409 of the same type , then @xmath450 carries any face @xmath451 of @xmath439 onto a face @xmath452 of @xmath12 so that @xmath453 have the same type as faces of @xmath409 and @xmath454 .
the diffeomorphisms @xmath455 will be called _
identification maps_. for example , every manifold with faces @xmath409 has a _ trivial partition _ such that two faces have the same type if and only if they coincide . given a partition @xmath443 on @xmath409 , we write @xmath456 for points
@xmath457 if there are faces @xmath458 of @xmath409 of the same type such that @xmath459 , @xmath460 , and @xmath461
. clearly , @xmath456 if and only if the faces @xmath462 have the same type and @xmath463
. then @xmath464 is an equivalence relation on @xmath409 .
the quotient topological space @xmath465 may not be a manifold .
for any set @xmath466 , we denote by @xmath467 the image of @xmath337 under the projection @xmath468 .
a smooth triangulation @xmath469 of @xmath409 _ fits a partition @xmath470 _ on @xmath409 if the identification map @xmath471 is a simplicial isomorphism for any faces @xmath449 of the same type .
[ partition ] for any partition @xmath443 on @xmath409 , there exists a smooth triangulation @xmath469 of @xmath409 which fits @xmath443 and projects to a triangulation , @xmath472 , of the quotient space @xmath473 .
we first construct by induction on @xmath413 a smooth triangulation @xmath474 of @xmath475 satisfying the following condition : all the identification maps between faces of @xmath409 of dimension @xmath476 are simplicial isomorphisms .
the case @xmath477 is obvious : we just take @xmath478 .
given @xmath479 , we construct @xmath480 as follows : pick one @xmath418-dimensional face of @xmath409 in each type and extend @xmath479 to the union of @xmath481 with these faces using the theory of smooth triangulations ( * ? ? ?
* section 10.6 ) .
the resulting triangulation of this union uniquely extends to a triangulation @xmath474 of @xmath475 satisfying the required condition .
set @xmath426 .
clearly , @xmath482 is a smooth triangulation of @xmath409 that fits @xmath443 .
let @xmath483 and @xmath484 be the first and second barycentric subdivisions of @xmath469 , respectively .
both @xmath483 and @xmath484 fit @xmath443 .
we claim that ( i ) the projection @xmath485 is injective on each simplex of @xmath483 and ( ii ) the images under @xmath11 of any two simplices of @xmath484 ( which by ( i ) are simplices ) meet along a common face .
thus , the triangulation @xmath484 of @xmath409 projects to a triangulation of @xmath473 and satisfies the conditions of the lemma . to prove ( i ) , consider a simplex @xmath486 of @xmath483 .
since all simplices of @xmath483 are faces of @xmath405-simplices , it suffices to consider the case where @xmath487 .
note that the restriction of @xmath488 to the interior of any face of @xmath409 is injective .
moreover , for any faces @xmath489 of @xmath409 , the restriction of @xmath11 to @xmath490 is injective .
therefore , to prove the injectivity of @xmath491 , it is enough to find a sequence of faces @xmath492 of @xmath409 , possibly with repetitions , such that @xmath493 .
let @xmath494 be the simplices of @xmath469 whose barycenters are the vertices of @xmath486 where @xmath495 for all @xmath496 .
let @xmath497 be the smallest face of @xmath409 containing @xmath498 .
the inclusions @xmath499 imply that @xmath500 for all @xmath496 .
note that @xmath501 since @xmath502 is a subcomplex of @xmath469 .
thus , @xmath503 . to prove ( ii ) , observe first that for any simplex @xmath504 of @xmath483 , the set @xmath505 is a subcomplex of @xmath483 . indeed , this set is equal to @xmath506 where @xmath449 run over all faces of @xmath409 of the same type .
since @xmath483 fits @xmath443 and both @xmath507 and @xmath508 are subcomplexes of @xmath509 , so are the sets @xmath510 , @xmath511 , and @xmath512 .
consider any simplices @xmath513 of the triangulation @xmath484 .
let @xmath514 and @xmath515 be simplices of @xmath483 containing @xmath516 and @xmath517 respectively .
set @xmath518 .
clearly , @xmath519 is the image of the set @xmath520 under @xmath11 . by the above
, the latter set is a subcomplex of @xmath514 .
therefore , @xmath519 is a subcomplex of the simplex @xmath521 .
we claim that @xmath522 is a face of the simplex @xmath523 .
this claim would imply that @xmath524 for a simplex @xmath525 of @xmath484 .
since @xmath526 , we can assume ( replacing if necessary @xmath525 by some @xmath527 ) that @xmath528 where @xmath529 is the barycentric subdivision of @xmath515 .
then @xmath530 is an intersection of two simplices of @xmath531 .
hence , it is a simplex of @xmath531 and a face of @xmath532 . by symmetry between @xmath516 and @xmath517 , the intersection @xmath533 is also a face of @xmath523 .
to prove the claim above , we need only to show that any subcomplex @xmath519 of an arbitrary simplex @xmath534 meets any simplex @xmath486 of the first barycentric subdivision @xmath535 along a face of @xmath486 .
clearly , @xmath536 is an intersection of two subcomplexes of @xmath535 and therefore a subcomplex of @xmath486 .
set @xmath537 and let @xmath538 be the faces of @xmath534 whose barycenters @xmath539 are the vertices of @xmath486 .
let @xmath496 be the largest integer such that @xmath540 .
since @xmath519 contains an interior point of @xmath498 and @xmath519 is a subcomplex of @xmath534 , we have @xmath541
. then @xmath519 contains the face @xmath542 of @xmath486 .
since @xmath536 is a subcomplex of @xmath486 not containing @xmath543 , we have @xmath544 . fix a topological space @xmath81 . given a partition @xmath443 on a manifold with faces
@xmath409 , we say that a continuous map @xmath545 is with @xmath470 if @xmath546 for any faces @xmath458 of @xmath409 of the same type .
every such @xmath73 is obtained by composing the projection @xmath547 with a continuous map @xmath548 .
an _ @xmath405-dimensional polychain _ or , shorter , an _
@xmath405-polychain _ in @xmath81 with @xmath406 is a quadruplet @xmath549 where @xmath409 is an oriented @xmath405-manifold with faces , @xmath550 is a partition on @xmath409 , @xmath551 is a map @xmath552 called the _ weight _ , and @xmath553 is a continuous map compatible with @xmath550 . by convention , for every @xmath406 , there is an _ empty @xmath405-polychain _ @xmath554 whose underlying @xmath405-manifold is the empty set . a _ diffeomorphism _ of @xmath405-polychains @xmath555 and @xmath556 in @xmath81 is a diffeomorphism @xmath557 such that * @xmath558 ; * faces @xmath449 of @xmath409 have the same type if and only the faces @xmath559 of @xmath560 have the same type and then @xmath561 ; * @xmath562 for any connected component @xmath563 of @xmath409 where @xmath564 denotes the degree of a diffeomorphism .
we say that @xmath405-polychains @xmath565 and @xmath566 in @xmath81 are _ diffeomorphic _ and we write @xmath567 if there exists a diffeomorphism of @xmath565 onto @xmath566 .
it is clear that @xmath568 is an equivalence relation . by definition ,
the diffeomorphism class of a polychain @xmath569 is preserved if one simultaneously inverts the orientation of a component of @xmath409 and multiplies the corresponding weight by @xmath570 . therefore the _ opposite _ polychain @xmath571 is diffeomorphic to @xmath572
. examples of polychains are provided by _
singular manifolds _ in @xmath81 , that is pairs of the form ( an oriented smooth compact manifold @xmath0 , a continuous map @xmath573 ) .
such a pair determines a polychain @xmath574 where @xmath575 is viewed as a manifold with faces as in section [ manifolds with faces ] , @xmath443 is the trivial partition , and @xmath576 is the constant function on @xmath577 . as explained below , polychains in @xmath81 may be also extracted from singular chains in @xmath81 .
thus , we can view polychains as common generalisations of singular manifolds and singular chains in which the role of source spaces is played by manifolds with faces .
a polychain @xmath578 in a topological space @xmath81 is if any distinct connected components of @xmath409 have different types with respect to @xmath443 and @xmath579 for any connected component @xmath563 of @xmath409 .
we define two transformations of an arbitrary polychain @xmath555 in @xmath81 whose composition turns @xmath580 into a reduced polychain . to define the first transformation , pick a representative in each type of connected components of @xmath409 , and let @xmath581 be the union of these representatives . clearly , @xmath582 is a manifold with faces which we endow with orientation induced from that of @xmath409 . restricting @xmath550 and @xmath73 to @xmath582 we
obtain a partition @xmath583 on @xmath582 and a map @xmath584 compatible with @xmath583 .
we define a weight @xmath585 on @xmath582 by @xmath586 where @xmath563 is a component of @xmath409 lying in @xmath582 and @xmath587 runs over all components of @xmath409 of the same type as @xmath563 .
it is clear that @xmath588 is a polychain in @xmath81 whose distinct components have different types .
this polychain , denoted @xmath589 , is determined by @xmath565 uniquely up to diffeomorphism .
the second transformation of a polychain @xmath555 removes from @xmath409 all connected components with zero weight and restricts @xmath590 to the remaining manifold with faces .
the resulting polychain is denoted @xmath591 .
the two - step operation @xmath592 transforms an arbitrary polychain into a reduced polychain defined uniquely up to diffeomorphism .
it is clear that a polychain @xmath580 is reduced if and only if @xmath593 .
the _ boundary _ of an @xmath405-polychain @xmath594 in a topological space @xmath81 is the @xmath424-polychain @xmath595 in @xmath81 defined as follows . * the manifold with faces @xmath596 is the disjoint union of all principal faces of @xmath409 endowed with orientation induced from that of @xmath409 ( see the introduction for our orientation conventions ) .
* let @xmath597 be the natural map identifying each component of @xmath598 with its copy in @xmath409 .
two faces @xmath449 of @xmath598 have the same type if the faces @xmath599 of @xmath409 have the same type and @xmath600 . * for any connected component @xmath601 of @xmath598 , we set @xmath602 where @xmath603 is the connected component of @xmath409 containing the principal face @xmath604 . *
we set @xmath605 .
the boundary of a polychain is well defined up to diffeomorphism , and diffeomorphic polychains have diffeomorphic boundaries . the _ reduced boundary _
@xmath606 of a polychain @xmath580 is defined by @xmath607 . for any polychain @xmath580 in @xmath81 , @xmath608 and @xmath609 .
the first identity is clear .
the second identity can be deduced from the first one as follows : @xmath610 the _ disjoint union _ of two @xmath405-polychains @xmath611 in @xmath81 is defined in the obvious way and is denoted @xmath612 .
clearly , @xmath613 so that @xmath614 . for @xmath138 and a polychain @xmath594 in @xmath81 , set @xmath615 .
clearly , @xmath616 so that @xmath617 . note that the polychain @xmath618 is diffeomorphic to the polychain @xmath619 opposite to @xmath580 .
though the diffeomorphism classes of @xmath405-polychains in a topological space @xmath81 may be added and multiplied by elements of @xmath95 , they do not form a module because the distributivity relation @xmath620 fails .
also , it is natural to throw in relations identifying @xmath580 with @xmath621 for all @xmath580 . quotienting the set of diffeomorphism classes of @xmath405-polychains in @xmath81 by all these relations , we obtain the _ @xmath95-module of @xmath405-polychains _ in @xmath81 . these modules together with the boundary maps induced by @xmath622
form the _ face chain complex _ of @xmath81 whose homology is the _ face homology _ of @xmath81 . however , for the purposes of this paper , we prefer the following more direct definition of face homology .
we say that @xmath405-polychains @xmath623 and @xmath624 in @xmath81 are _ homologous _ , and write @xmath625 , if there exist @xmath626-polychains @xmath627 in @xmath81 such that @xmath628 clearly , the homology relation @xmath629 is an equivalence relation ( weaker than the diffeomorphism relation @xmath630 ) .
the homology class of an @xmath405-polychain @xmath580 in @xmath81 is denoted by @xmath631 .
note that @xmath632 .
if @xmath623 and @xmath624 are homologous , then @xmath633 .
a polychain @xmath634 is a _ polycycle _ if @xmath635 .
a polychain homologous to a polycycle is itself a polycycle . in particular , if @xmath580 is a polycycle , then so is @xmath636 and vice versa
let @xmath637 be the set of homology classes of @xmath405-polycycles in @xmath81 .
note that the disjoint union of polycycles is a polycycle , and multiplication of polycycles by elements of @xmath95 yield polycycles .
[ additivityandrescaling ] disjoint union of polycycles together with multiplication of polycycles by elements of @xmath95 turn @xmath638 into a module ( over @xmath95 ) .
clearly , the disjoint union of polychains is compatible with @xmath639 and induces a binary operation in @xmath638 .
this operation is associative and commutative with @xmath554 representing the zero element .
thus , @xmath640 is an abelian monoid . to prove that @xmath638 is a group
, we use the cylinder construction on polychains . consider an @xmath405-polychain @xmath594 in @xmath81 .
we define the @xmath641 as follows .
set @xmath642 where @xmath643 $ ] is viewed as a manifold with faces @xmath644 , @xmath645 , @xmath646 and endow @xmath647 with the product orientation .
two faces @xmath648 , @xmath649 of @xmath650 are of the same type if @xmath449 are faces of @xmath409 of the same type and @xmath651 is any face of @xmath644 ; then @xmath652 . by definition ,
@xmath653 for any connected component @xmath563 of @xmath409 , and @xmath654 is the composition of the cartesian projection @xmath655 with @xmath656 .
it follows from the definitions that @xmath657 therefore @xmath658 if @xmath580 is a polycycle , this gives @xmath659 .
therefore @xmath660 .
we conclude that @xmath638 is an abelian group . given two homologous @xmath405-polycycles @xmath623 and @xmath624 in @xmath81 , pick @xmath626-polychains @xmath627 in @xmath81 such that @xmath661 .
then , for any @xmath138 , @xmath662 for each @xmath663 , @xmath664 we deduce that @xmath665 .
thus , the multiplication by @xmath138 induces a well defined map @xmath666 .
the axioms of a @xmath95-module are straightforward except the linearity in @xmath189 .
the latter is a consequence of the following fact : if @xmath667 and @xmath668 are @xmath405-polycycles in @xmath81 ( with the same @xmath669 ) , then the @xmath405-polychain @xmath670 is a polycycle homologous to @xmath612 . to see this ,
consider the cylinder polychain @xmath671 ( as defined above ) and modify its partition by additionally declaring that , for any face @xmath439 of @xmath409 , the faces @xmath672 and @xmath673 of @xmath674 have the same type and the corresponding identification map is the identity map .
this gives an @xmath626-polychain @xmath675 such that @xmath676 therefore @xmath677 hence , @xmath580 is a polycycle homologous to @xmath678 .
we call @xmath638 the _ @xmath405-th face homology of @xmath81 _ ( with coefficients in @xmath95 ) .
the face homology extends to a functor from the category of topological spaces to the category of modules : a continuous map @xmath679 induces a linear map @xmath680 carrying the homology class of a polycycle @xmath681 in @xmath81 to the homology class of the polycycle @xmath682 in @xmath683 .
a of a polychain @xmath634 in a topological space @xmath81 is a family of polychains @xmath684 with the same @xmath685 such that @xmath686 is a ( continuous ) homotopy of @xmath687 .
note that , by the definition of a polychain , the map @xmath688 must be compatible with @xmath443 for all @xmath689 .
[ homotimplieshomol ] if @xmath690 is a deformation of a polycycle @xmath580 , then @xmath691 is a polycycle homologous to @xmath692 .
equality @xmath693 is a direct consequence of the assumption @xmath694 .
consider the cylinder polychain @xmath641 associated with @xmath634 in the proof of lemma [ additivityandrescaling ] .
let @xmath695 be the map determined by the homotopy @xmath696 of @xmath73 .
then @xmath697 is a polychain such that @xmath698 this implies that @xmath699 .
[ homotimplieshomol+ ] let @xmath351 be topological spaces .
if maps @xmath700 are homotopic , then @xmath701 .
pick a homotopy @xmath702 between @xmath703 and @xmath704 . for any polycycle @xmath634 in @xmath81
, we have a deformation @xmath705 relating the polycycles @xmath706 and @xmath707 .
lemma [ homotimplieshomol ] implies that these polycycles are homologous .
hence , @xmath708 .
lemma [ homotimplieshomol+ ] implies , in particular , that a homotopy equivalence between topological spaces induces an isomorphism of their face homology .
the product @xmath709 of two manifolds with faces @xmath409 and @xmath337 can be viewed as a manifold with faces in the obvious way .
the faces of @xmath709 are the products @xmath710 where @xmath439 runs over faces of @xmath409 and @xmath12 runs over faces of @xmath337 .
when @xmath409 and @xmath337 are oriented , we always provide @xmath709 with the product orientation .
this construction leads to a cross product in face homology as follows .
let @xmath81 and @xmath683 be topological spaces .
cross product _ of a @xmath102-polychain @xmath634 in @xmath81 and a @xmath711-polychain @xmath712 in @xmath683 is the @xmath713-polychain @xmath714 in @xmath715 . here
@xmath716 is the following partition on @xmath709 : for faces @xmath717 of @xmath409 and @xmath718 of @xmath337 , the faces @xmath719 and @xmath720 of @xmath709 have the same type if , and only if , @xmath439 has the same type as @xmath451 and @xmath12 has the same type as @xmath452 , and then @xmath721 the weight @xmath722 carries @xmath723 to @xmath724 for any connected components @xmath563 of @xmath409 and @xmath56 of @xmath337 .
we also define the _ reduced cross product _ of @xmath580 and @xmath675 by @xmath725 note that @xmath726 [ properties_cross ] ( i ) for any @xmath102-polychains @xmath611 in @xmath81 and any @xmath711-polychain @xmath675 in @xmath683 , @xmath727 ( ii ) for any @xmath102-polychain @xmath580 in @xmath81 and for any @xmath711-polychain @xmath675 in @xmath683 , @xmath728 clearly , @xmath729 so that @xmath730 which proves ( i ) .
we now prove ( ii ) .
let @xmath409 and @xmath337 be the oriented manifolds with faces underlying @xmath580 and @xmath675 respectively .
a principal face of @xmath709 has either the form @xmath731 , where @xmath601 is a principal face of @xmath409 and @xmath56 is a component of @xmath337 , or the form @xmath732 , where @xmath563 is a component of @xmath409 and @xmath733 is a principal face of @xmath337 .
the orientation of @xmath734 inherited from @xmath735 coincides with the product orientation of @xmath731 where @xmath736 inherits orientation from @xmath409 .
the orientation of @xmath737 inherited from @xmath735 differs from the product orientation of @xmath738 , where @xmath739 inherits orientation from @xmath337 , by the sign @xmath740 .
so @xmath741 therefore @xmath742 we conclude thanks to .
the cross product of polychains in @xmath81 and @xmath683 induces a bilinear map @xmath743 let @xmath580 be a polycycle in @xmath81 and @xmath675 be a polycycle in @xmath683 .
lemma [ properties_cross].(ii ) implies that @xmath744 is a polycycle .
this polycycle is the reduction of @xmath745 , and therefore @xmath745 also is a polycycle .
we claim that assigning to @xmath746 the homology class @xmath747 one obtains a well defined pairing .
let us prove the independence of the choice of @xmath580 in its homology class ( the second variable is treated similarly ) .
consider two homologous polycycles @xmath623 and @xmath624 in @xmath81 , and let @xmath748 be polychains in @xmath81 such that @xmath749 lemma [ properties_cross].(i ) implies that @xmath750 for @xmath663 , formula gives @xmath751 . since @xmath752 , lemma [ properties_cross].(ii ) gives @xmath753 .
therefore @xmath754 .
the linearity of in the first variable follows from lemma [ properties_cross].(i ) and the equality @xmath755 for all @xmath138 .
the linearity in the second variable is proved similarly .
a polychain derived from a singular manifold @xmath756 ( see section [ polychains ] ) is a polycycle if and only if @xmath757 . the oriented bordism classes of @xmath405-dimensional singular manifolds @xmath758 with @xmath759 form an abelian group @xmath760 , called the _ @xmath405-dimensional oriented bordism group _ of @xmath81 . treating singular manifolds as polychains , we obtain an additive map @xmath761 .
as explained in remark [ rerererer].2 below , this map is not surjective for some @xmath762 , and @xmath763 .
thus , some face homology classes over @xmath764 are not representable by singular manifolds . \2 . for a topological pair @xmath765 and an integer @xmath406
, we define the @xmath405-th _ relative face homology _ @xmath766 as follows .
given @xmath405-polychains @xmath767 in @xmath81 , we write @xmath768 , if there exist @xmath626-polychains @xmath627 in @xmath81 and @xmath405-polychains @xmath769 in @xmath683 such that @xmath770 where @xmath771 is the inclusion map .
an @xmath405-polychain @xmath580 in @xmath81 is a _ polycycle relative to @xmath683 _ if @xmath606 is the image of an @xmath424-polychain in @xmath683 under @xmath772 .
set @xmath773 the properties of the face homology of topological spaces stated above directly extend to the face homology of topological pairs .
in this section , we construct two natural transformations @xmath774:\widetilde { h}_\ast\to h_\ast$ ] and @xmath775 relating face homology to singular homology .
we begin by fixing notation and terminology on singular homology and triangulations . for an integer @xmath406 ,
the symbol @xmath776 denotes the standard @xmath405-simplex that is the convex hull of the standard basis @xmath777 of @xmath778 .
we endow @xmath776 with orientation induced by the order of its vertices , i.e@xmath134 the orientation represented by the basis @xmath779 in the tangent space of @xmath776 at any point .
each subset @xmath780 of @xmath781 with @xmath782 and @xmath783 determines an affine map @xmath784 carrying the vertices @xmath785 of @xmath786 to the vertices @xmath787 of @xmath776 , respectively ; the image of the map @xmath788 is the _ combinatorial face _ of @xmath776 corresponding to @xmath7 .
a in a topological space @xmath81 is a continuous map @xmath789 .
a in @xmath81 is a finite formal linear combination of singular @xmath405-simplices with coefficients in @xmath95 .
the _ boundary _ of a singular @xmath405-simplex @xmath790 is the singular @xmath424-chain @xmath791 the boundary of singular simplices extends to singular chains by linearity .
the modules of singular chains together with the boundary homomorphisms form the _ singular chain complex _
@xmath792 of @xmath81 .
its homology is the _
singular homology _
@xmath793 of @xmath81 ( with coefficients in @xmath95 ) .
consider an @xmath405-dimensional oriented manifold with faces @xmath409 .
each weight @xmath794 determines a homology class @xmath795=\sum_{c } u(c ) \big[c\big ] \
\in \bigoplus_c h_n ( c,\partial c ) = h_n ( k,\partial k)\ ] ] where @xmath563 runs over all connected components of @xmath409 and @xmath796\in h_n ( c,\partial c)$ ] is the fundamental class of @xmath563 .
we say that a partition @xmath443 on @xmath409 is with @xmath797 if for any principal face @xmath601 of @xmath409 , @xmath798 where @xmath733 runs over all ( principal ) faces of @xmath409 of the same type as @xmath601 and @xmath799 is the connected component of @xmath409 containing @xmath733 .
[ partit ] if a partition @xmath443 on @xmath409 is compatible with a weight @xmath794 , then there is a unique homology class @xmath800\in h_n(k_{\varphi})$ ] whose image in @xmath801 is equal to the image of @xmath802 $ ] under the map @xmath803 induced by the projection @xmath804 .
consider the commutative diagram @xmath805 \ar[d]^-{\pi _ * } & h_n(k ) \ar[r ] \ar[d]^-{\pi _ * } & h_n(k,\partial k ) \ar[r]^-{\partial _ * } \ar[d]^-{\pi _ * } & h_{n-1}(\partial k ) \ar[d]^-{\pi_*}\\ h_n\left((\partial k)_\varphi\right ) \ar[r ] & h_n\left(k_\varphi\right ) \ar[r ] & h_n\left(k_\varphi,(\partial k)_\varphi \right ) \ar[r]^-{\partial _ * } & h_{n-1}\left((\partial k)_\varphi\right ) , } \ ] ] where the vertical maps are induced by the projection @xmath488 and each row is a part of the long exact sequence of a topological pair .
we have @xmath806 since @xmath807 is an @xmath424-dimensional polyhedron .
hence , it is enough to show that @xmath808)=0 \in h_{n-1}\left((\partial k)_\varphi\right).\ ] ] consider the commutative square @xmath809)\in \!\!\ ! \!\!\!\!\!\ ! \!\!\ ! \!\!\ ! \!\!\ ! & h_{n-1}(\partial k ) \ar[r]^{j } \ar[d]^-{\pi _ * } & h_{n-1}\left(\partial k , k_{n-2}\right ) \ar[d]^-{\pi _ * } \\ & h_{n-1}\left((\partial k)_\varphi\right ) \ar[r]^-{j_\varphi } & h_{n-1}\left((\partial k)_\varphi , ( k_{n-2})_\varphi\right ) } \ ] ] where @xmath810 and @xmath811 are the inclusion homomorphisms . since @xmath812 is an @xmath813-dimensional polyhedron , @xmath814 and it suffices to prove that @xmath815)=0 $ ] or , equivalently , that @xmath816)=0 $ ] .
we have @xmath817 ) = \sum_c u(c ) \partial_*([c ] ) = \sum_c u(c ) [ \partial c]\ ] ] where the sum runs over the connected components @xmath563 of @xmath409 . then @xmath818)=\sum_{c } u(c ) \sum_{p \subset c } [ p ] = \sum_{p } u(k^p)\ , [ p]\ ] ] where @xmath601 runs over all principal faces of @xmath409 and @xmath603 is the connected component of @xmath409 containing @xmath601 .
pick a face @xmath819 in each type @xmath496 of principal faces of @xmath409 . then @xmath820 ) & = & \sum_{p } u(k^p)\ , \pi_*([p])\\ & = & \sum_{i } \sum_{q\in i } u({k^q})\ , \pi_*([q ] ) \\ & = & \sum_{i } \sum_{q \in i } u({k^q})\ , \pi _ * \big(\deg(\varphi_{p_i , q})\cdot ( \varphi_{p_i , q})_{*}([p_i ] ) \big)\\ & = & \sum_{i } \big(\sum_{q \in i } u({k^q } ) \deg(\varphi_{p_i , q})\big ) \ , \pi _ * \left([p_i ] \right)=0\end{aligned}\ ] ] where at the last step we use the compatibility condition .
it follows from the definitions that a polychain @xmath821 in a topological space @xmath81 is a polycycle if and only if @xmath797 and @xmath443 are compatible in the sense of .
therefore , given an @xmath405-polycycle @xmath594 in @xmath81 , lemma [ partit ] gives the homology class @xmath800\in h_n(k_\varphi)$ ] .
since the map @xmath822 is compatible with @xmath443 , it induces a continuous map @xmath823 denoted by @xmath824 .
we define @xmath825 = ( \kappa_{\varphi})_\ast ( [ k_{\varphi},u])\in h_n ( x).\ ] ] the homology class @xmath826 $ ] can be represented by explicit singular cycles which are best described in terms of locally ordered triangulations .
a _ local order _ on a triangulation @xmath469 of a topological space is a binary relation on the set of vertices of @xmath469 which restricts to a total order on the set of vertices of any simplex of @xmath469 .
for example , any total order on the set of vertices of @xmath469 is a local order on @xmath469 .
a triangulation endowed with a local order is _ locally ordered_. we say that a locally ordered smooth triangulation @xmath469 of @xmath409 _ fits the partition @xmath470 _ if , for any faces @xmath449 of @xmath409 of the same type , the identification map @xmath471 is a simplicial isomorphism preserving the local order on the vertices .
to construct such a locally ordered triangulation one can take a triangulation @xmath469 of @xmath409 provided by lemma [ partition ] and lift an arbitrary total order @xmath827 on the set of vertices of @xmath472 to @xmath469 .
more precisely , denote by @xmath485 the canonical projection and , for any vertices @xmath828 , declare that @xmath829 if @xmath830 .
since any simplex of @xmath469 projects isomorphically onto a simplex of @xmath472 , this gives a local order on @xmath469 which , obviously , fits @xmath443 .
pick an arbitrary locally ordered smooth triangulation @xmath469 of @xmath409 which fits @xmath443 .
each @xmath418-simplex @xmath534 of @xmath469 with @xmath413 determines a singular simplex in @xmath409 denoted by @xmath831 and obtained as the composition of the affine isomorphism @xmath832 preserving the order of the vertices with the inclusion @xmath833 .
we define the _ fundamental @xmath405-chain _
@xmath834 where @xmath534 runs over all @xmath405-simplices of @xmath469 , @xmath835 is the connected component of @xmath409 containing @xmath534 , @xmath836 if the orientation of @xmath534 induced by that of @xmath409 is compatible with the order of the vertices of @xmath534 and @xmath837 otherwise . clearly
the image of @xmath838 in @xmath839 is a relative @xmath405-cycle representing @xmath802 $ ] .
projecting @xmath838 to @xmath473 , we obtain a singular @xmath405-chain @xmath840 in @xmath841 .
the compatibility of @xmath443 and @xmath797 implies that @xmath842 is an @xmath405-cycle .
therefore the homology class @xmath843\in h_n(k_\varphi)$ ] satisfies the requirements of lemma [ partit ] so that @xmath844=[\sigma_\varphi]$ ] .
it follows that @xmath826 $ ] is represented by the singular @xmath405-cycle @xmath845 [ from_sc_to_pcdebut ] the formula @xmath846 $ ] defines a linear map @xmath774:\widetilde{h}_n(x ) \to { h}_n(x)$ ] .
moreover , @xmath774 $ ] is a natural transformation from @xmath847 to @xmath848 .
it follows from the definitions that @xmath826\in h_n(x)$ ] depends only on the diffeomorphism class of @xmath580 , that @xmath826=[{\operatorname{red}}({\mathscr{k}})]$ ] , and that @xmath849=k[{\mathscr{k}}]$ ] for any @xmath138 . moreover , @xmath850=[{\mathscr{k}}_1]+ [ { \mathscr{k}}_2]$ ] for any @xmath405-polycycles @xmath767 in @xmath81 .
therefore , to prove the first claim of the lemma , it is enough to show that @xmath851=0 $ ] for any @xmath626-polychain @xmath852 in @xmath81 . for this , pick a locally ordered smooth triangulation @xmath469 of @xmath337 that fits @xmath853 and consider the singular chain @xmath854 here @xmath534 runs over all @xmath626-dimensional simplices of @xmath469 , @xmath855 is the sign determined by the orientation of @xmath337 and the order of the vertices of @xmath534 , and @xmath856 is the component of @xmath337 containing @xmath534 .
projecting @xmath838 to @xmath857 we obtain a singular chain @xmath858 in @xmath859 .
next we consider the @xmath405-polycycle @xmath860 .
the triangulation @xmath469 of @xmath337 induces a triangulation @xmath861 of @xmath862 .
the local order on the set of vertices of @xmath469 restricts to a local order on the set of vertices of @xmath861 .
consider the fundamental @xmath405-chain @xmath863 in @xmath862 as defined before the statement of the lemma .
projecting @xmath486 to the quotient space @xmath864 we obtain a singular @xmath405-cycle @xmath865 representing @xmath866 \in h_n\big((l^\partial)_{\psi^\partial}\big)$ ] .
the natural map @xmath867 induces a map @xmath868 carrying @xmath865 to @xmath869 . by definition ,
therefore @xmath871 where @xmath872 is the map induced by @xmath873 .
hence , @xmath874 = \big((\lambda^\partial)_{\psi^\partial } \big ) _ * ( [ \tau_{\psi^\partial } ] ) = ( \lambda_\psi \iota_\psi ) _ * ( [ \tau_{\psi^\partial } ] ) = ( \lambda_\psi ) _ * ( [ \partial \sigma_{\psi } ] ) = 0.\ ] ] to prove the second claim of the lemma , consider a continuous map @xmath679 . for any @xmath405-polycycle @xmath634 in @xmath81
, we have @xmath875\right ) = f_*(\kappa_{\varphi})_\ast ( [ k_{\varphi},u ] ) = ( f \kappa_{\varphi})_\ast ( [ k_{\varphi},u ] ) = \big ( ( f \kappa)_{\varphi}\big)_\ast ( [ k_{\varphi},u ] ) = [ f_*({\mathscr{k}})]\ ] ] since @xmath876 by definition .
let @xmath81 be a topological space and let @xmath406 be an integer .
we associate with each singular @xmath405-chain @xmath838 in @xmath81 an @xmath405-polychain @xmath878 in @xmath81 .
pick an expansion @xmath879 , where @xmath496 runs over a finite set of indices , @xmath880 , and @xmath881 are singular @xmath405-simplices in @xmath81 .
let @xmath409 be the manifold with faces obtained as a disjoint union of copies @xmath882 of @xmath776 numerated by all @xmath883 .
we define a partition @xmath470 on @xmath409 as follows : a face @xmath439 of @xmath882 corresponding to a set @xmath884 and a face @xmath451 of @xmath885 corresponding to a set @xmath886 are declared to be of the same type if , and only if , @xmath7 and @xmath887 have the same cardinality , @xmath888 , and @xmath889 ( where @xmath890 are the maps defined in section [ singular_homology ] ) .
then we set @xmath891 .
clearly , the map @xmath892 is compatible with @xmath443 .
we define a weight @xmath893 by @xmath894 for all @xmath496 .
the tuple @xmath895 is an @xmath405-polychain in @xmath81 depending on the choice of the expansion @xmath896 .
however , the polychain @xmath897 does not depend on this choice .
indeed , any two expansions of @xmath838 may be related by the following operations : replacement of @xmath898 by @xmath899 for any @xmath900 and any singular @xmath405-simplex @xmath901 in @xmath81 ; addition of a term @xmath902 for an arbitrary singular @xmath405-simplex @xmath901 in @xmath81 ; the inverse operations .
it is easy to see that @xmath878 is preserved under these transformations . by definition , if @xmath903 , then @xmath904 . the face homology class @xmath905 of the polychain @xmath878 will be denoted by @xmath906 .
[ mapp ] if @xmath838 is a cycle , then @xmath907 is a polycycle .
the formula @xmath908 \mapsto { \left\langle}\sigma { \right\rangle}$ ] , applied to singular @xmath405-cycles in @xmath81 , defines a linear map @xmath909 .
moreover , @xmath910 is a natural transformation from @xmath848 to @xmath847 .
we check first that for any singular @xmath405-chain @xmath838 in @xmath81 , @xmath911 pick an expansion @xmath879 such that the simplices @xmath912 are pairwise distinct and @xmath913 for all @xmath496 .
then the associated polychain @xmath895 is reduced and @xmath914 .
a connected component @xmath601 of @xmath915 is nothing but a principal face of @xmath916 for some @xmath917 corresponding to the complement of a singleton @xmath918 . by the definition of @xmath919
, we have @xmath920 .
we compute @xmath921 as described in section [ polychains ] .
pick a representative @xmath601 for each type of connected components of @xmath598 , and let @xmath922 be the union of these representatives . restricting @xmath923 and @xmath924 to @xmath925 we obtain a partition @xmath926 on @xmath925 and a compatible map @xmath927 .
the weight @xmath928 on @xmath925 is evaluated on each component @xmath601 of @xmath925 by @xmath929 where @xmath733 runs over all components of @xmath598 of the same type as @xmath601 .
note that @xmath930 is the total coefficient of the singular simplex @xmath931 in @xmath932 .
also , @xmath933 is the degree of @xmath934 ( recall that @xmath935 is oriented as in section [ singular_homology ] while @xmath936 inherits orientation from @xmath882 where @xmath917 ) .
we conclude that the polychain @xmath937 consists of @xmath938 and eventually several connected components of weight zero .
hence @xmath939 this proves .
the first assertion of the lemma follows .
next we claim that @xmath940 for any singular @xmath405-cycles @xmath941 in @xmath81 . to see this , pick expansions @xmath942 , @xmath943 and let @xmath634 , @xmath712 be the associated polychains , respectively .
consider the cylinder polychain @xmath944 ( as defined in the proof of lemma [ additivityandrescaling ] ) and modify its partition by additionally declaring that for any face @xmath439 of @xmath916 corresponding to @xmath945 and for any face @xmath12 of @xmath946 corresponding to @xmath947 such that @xmath948 , the faces @xmath949 and @xmath950 of @xmath951 are of the same type , and the corresponding identification map is @xmath952 .
the resulting @xmath626-polychain , @xmath953 , in @xmath81 satisfies @xmath954 where @xmath955 is the polychain associated with the expansion @xmath956 of @xmath957 .
hence , @xmath958 and our claim follows . if @xmath95 has no zero - divisors , then @xmath959 for any singular @xmath405-chain @xmath838 in @xmath81 and any non - zero @xmath138 .
for an arbitrary @xmath95 and all @xmath138 , we have @xmath960 equalities imply that the formula @xmath908 \mapsto { \left\langle}{\mathscr{p}}(\sigma ) { \right\rangle}$ ] defines a linear map @xmath961 . to prove the last claim of the lemma ,
consider a continuous map @xmath679 .
let @xmath838 be a singular @xmath405-cycle in @xmath81 , and let @xmath634 be the @xmath405-polycycle associated to an expansion @xmath962 of @xmath838 .
the @xmath405-polycycle associated to the expansion @xmath963 of @xmath964 has the form @xmath965 and differs from @xmath966 only in the partition .
modifying appropriately the partition of the cylinder polychain @xmath967 , we obtain an @xmath626-polychain @xmath968 in @xmath81 such that @xmath969 we deduce that @xmath970 and @xmath971 the next theorem implies that @xmath972 is canonically isomorphic to a direct summand of @xmath638 .
[ mainonpolychains ] we have @xmath973\circ { \left\langle}-{\right\rangle}= { \operatorname{id}}:h_n(x)\to h_n(x)$ ] .
let @xmath974 be a singular @xmath405-cycle in @xmath81 and let @xmath975 be the corresponding reduced @xmath405-polycycle
. then @xmath976 { \right\rangle}\right ] = [ { \mathscr{p}}(\sigma)]= ( \kappa_{\varphi})_*\left([k_{\varphi},u]\right ) = \big[\sum_i k_i \sigma_i\big]= [ \sigma ] \in h_n(x).\ ] ] here the third equality is obtained by considering the tautological locally ordered smooth triangulation @xmath469 of @xmath409 and the corresponding fundamental @xmath405-chain @xmath977 ( see the paragraph preceding lemma [ from_sc_to_pcdebut ] ) .
the following lemma shows that the transformation @xmath774 : \widetilde h_*\to h_\ast$ ] carries the cross product @xmath978 in face homology to the standard cross product @xmath978 in singular homology .
[ twocrossproducts ] for any topological spaces @xmath81 , @xmath683 , the following diagram commutes : @xmath979 ^ -{\times } \ar[d]_{[-]\times [ - ] } & \widetilde h_*(x \times y ) \ar[d]^{[-]}\\
h_*(x ) \times
h_*(y ) \ar[r]^-{\times } & h_*(x \times y ) . } \ ] ] we recall the definition of the map @xmath980 and then prove lemma [ twocrossproducts ] .
fix integers @xmath981 .
any @xmath102-element subset @xmath982 of @xmath983 determines non - decreasing maps @xmath984 such that @xmath985 and for any @xmath986 , @xmath987 let @xmath988 be the convex hull of the set @xmath989 . [ ez ] the set @xmath990 is an embedded @xmath713-simplex with vertices @xmath991 for @xmath992 .
moreover , the simplices @xmath993 and their faces form a triangulation of @xmath994 . this lemma is well known but we give a proof for completeness . for @xmath406 ,
denote by @xmath995 the affine space formed by the points of @xmath778 with sum of coordinates @xmath300 .
the basis @xmath996 of @xmath778 is an affine basis of @xmath995 and @xmath997 .
consider the basis @xmath998 of the vector space underlying @xmath999 and the basis @xmath1000 of the vector space underlying @xmath1001 .
then @xmath1002 is a basis of the vector space underlying the product affine space @xmath1003 .
recall that a _ @xmath1004-shuffle _ is a permutation @xmath1005 of @xmath983 such that @xmath1006 any @xmath102-element subset @xmath982 of @xmath983 determines a unique @xmath1004-shuffle @xmath1005 such that @xmath1007 .
the first claim of the lemma follows from the fact that the vector basis @xmath1008 underlies the set of vertices of @xmath1009 : @xmath1010 ^ -{v_{s^{-1}(1 ) } } & ( e_{\alpha(1)},e_{\beta(1 ) } ) \ar@{|->}[r]^-{v_{s^{-1}(2 ) } } & \ \cdots \ \ar@{|->}[r]^-{v_{s^{-1}(p+q ) } } & ( e_{\alpha(p+q)},e_{\beta(p+q ) } ) . } \ ] ] to prove the second claim , observe that given @xmath1011 affinely independent points @xmath1012 in an @xmath405-dimensional affine space , an arbitrary point @xmath1013 of this space ( with @xmath1014 ) belongs to the affine simplex spanned by @xmath1012 if and only if @xmath1015 . therefore any point @xmath1016 expands uniquely as @xmath1017 with @xmath1018 by the same observation and , the inclusion @xmath1019 for a @xmath102-element subset @xmath982 of @xmath1020 holds if and only if @xmath1021 where @xmath1005 is the @xmath1004-shuffle determined by @xmath982 . therefore @xmath994 is the union of the simplices @xmath993 , and any two of these simplices meet along a common face . for each @xmath102-element subset @xmath1022 , we turn @xmath990 into a singular simplex in @xmath994 by sending the ordered vertices @xmath1023 of @xmath1024 to @xmath1025 respectively . summing up over all such @xmath982
we obtain a singular chain @xmath1026 where @xmath1027 is the sign comparing the orientation in @xmath990 determined by the order of its vertices with the product orientation in @xmath1028 .
then the _ eilenberg zilber chain map _ @xmath1029 is defined by @xmath1030 for any singular simplices @xmath1031 and @xmath1032 . here
@xmath1033 is the chain map @xmath1034 induced by @xmath1035 . the _ cross product _ of singular homology classes @xmath1036 and
@xmath1037 is defined by taking any cycles @xmath1038 and @xmath1039 representing @xmath209 and @xmath1040 respectively , and letting @xmath1041 be the homology class of @xmath1042 .
let @xmath634 be a @xmath102-polycycle in @xmath81 and let @xmath712 be a @xmath711-polycycle in @xmath683 .
we must prove that @xmath1043_bis } [ { \mathscr{k}}\times { \mathscr{l } } ] = [ { \mathscr{k } } ] \times [ { \mathscr{l } } ] \in h_{p+q}(x \times y).\ ] ] fix a locally ordered smooth triangulation @xmath469 of @xmath409 which fits @xmath443 and consider the fundamental @xmath102-chain @xmath1044 where @xmath496 runs over @xmath102-simplices of @xmath469 , @xmath1045 is the smooth singular simplex determined by @xmath496 , @xmath1046 is the sign comparing the orientation induced by the order of the vertices of @xmath496 to the orientation of @xmath409 , and @xmath1047 is the connected component of @xmath409 containing @xmath496 . then @xmath825 = \big[\sum_i \ , \varepsilon_i \ , u(k^i ) \ ,
\left(\kappa \circ \sigma_i\right ) \big ] \in h_p(x)\ ] ] where the square brackets on the right - hand side stand for the homology class of a singular cycle . similarly , fixing a locally ordered smooth triangulation @xmath93 of @xmath337 which fits @xmath853 , we obtain @xmath1048 = \big[\sum_j \ , \varepsilon_j\ , v(l^j ) \ , \left(\lambda \circ \tau_j\right ) \big ] \in h_q(y)\ ] ] where @xmath810 runs over @xmath711-simplices of @xmath93 , @xmath1049 is the smooth singular simplex determined by @xmath810 , @xmath1050 is the sign comparing the orientation induced by the order of the vertices of @xmath810 to the orientation of @xmath337 , and @xmath1051 is the component of @xmath337 containing @xmath810 . by the definition of the cross product in singular homology , @xmath1052 \times [ { \mathscr{l } } ] & = & \big[\sum_{i , j } \ , \varepsilon_i\ , \varepsilon_j\ , u(k^i ) \ , v(l^j ) \ , ez(\kappa\sigma_i \otimes \lambda\tau_j ) \big ] \\ \label{[k]x[l]_bis } & = & \big [ \sum_{i , j , s } \ ,
\varepsilon_i\ , \varepsilon_j\ , \varepsilon_s \ , u(k^i ) \ , v(l^j ) \,\ , ( \kappa \sigma_i \times \lambda\tau_j ) \omega_s\big]\end{aligned}\ ] ] where @xmath982 runs over @xmath102-element subsets of @xmath1053 .
for any simplices @xmath136 as above , we push forward via @xmath1054 the triangulation of @xmath1055 provided by lemma [ ez ] to a triangulation of @xmath1056 .
this gives a smooth triangulation @xmath1057 of @xmath735 .
the set of vertices @xmath1058 of @xmath1057 is the cartesian product of the sets of vertices @xmath1059 and @xmath1060 of @xmath469 and @xmath93 , respectively ; we endow @xmath1058 with the product of the binary relations on @xmath1059 and @xmath1060 determined by the local orders on @xmath469 and @xmath93 .
this defines a local order on @xmath1057 which fits the partition @xmath716 .
the simplices of @xmath1057 can be identified with the triples @xmath1061 as above , and the corresponding singular simplices in @xmath709 are the maps @xmath1062 .
( here we use the fact that the order of the vertices of @xmath990 is given by the product binary relation on the set of vertices of @xmath1055 determined by the natural orders on the sets of vertices of @xmath1063 and @xmath1064 . )
let @xmath1065 be the sign comparing the orientation of the simplex @xmath1061 induced by the order of the vertices to the product orientation of @xmath709 .
let @xmath1066 be the component of @xmath709 containing the simplex @xmath1061 .
then @xmath1043 } [ { \mathscr{k}}\times { \mathscr{l } } ] = \big[\sum_{i , j , s } \varepsilon_{i , j , s}\cdot ( u \times v)\left((k\times l)^{i , j , s}\right ) \cdot \big((\kappa \times \lambda)\circ ( \sigma_i \times \tau_j ) \omega_s\big ) \big].\ ] ] clearly , @xmath1067 note that @xmath1068 since @xmath1069 is the degree of the diffeomorphism @xmath1070 with respect to the product orientation in @xmath994 and the product orientation in @xmath735 restricted to @xmath1071 . comparing to , we obtain . \1 . though we shall not need it in the sequel , note that the sign @xmath1027 in can be computed explicitly : @xmath1072 where @xmath1073 is the number of pairs @xmath1074 with @xmath1075 and @xmath1076 .
indeed , in the notation introduced in the proof of lemma [ ez ] , the orientation of @xmath990 determined by the sequence is represented by the @xmath713-vector @xmath1077 where @xmath1005 is the @xmath1004-shuffle associated with @xmath982 and @xmath435 is the number of inversions in @xmath1005 . therefore @xmath1078 and it remains to observe that @xmath1079 .
the definition of @xmath1073 may be also reformulated in terms of the maps @xmath1080 and @xmath1081 .
namely , @xmath1073 is the number of pairs @xmath1074 such that @xmath1082 and @xmath1083 .
this implies the following formula for @xmath1073 used , for example , in ( * ? ? ?
* section 4(b ) ) : @xmath1084 \2 . by a celebrated result of r. thom
, there are topological spaces @xmath81 and integers @xmath1085 such that some @xmath405-dimensional singular homology classes of @xmath81 with coefficients in @xmath763 are not realizable by closed singular manifolds . for such @xmath81 and @xmath405 , the natural map from the @xmath405-dimensional oriented bordism group @xmath760 to @xmath1086 , carrying a closed singular manifold @xmath1087 to @xmath1088)$ ] ,
is not surjective .
this map splits as a composition of the map @xmath1089 described in remark [ singsing ] with the surjective map @xmath774:\widetilde { h}_n ( x)\to h_n ( x)$ ] .
therefore , for such @xmath81 and @xmath405 , the map @xmath1089 is not surjective .
the face homology is difficult to compute . as a consequence
, the authors do not know whether the transformation @xmath774:\widetilde { h}_\ast\to h_\ast$ ] is injective , and equivalently , whether @xmath775 is surjective .
( we believe they are not . ) \4 .
the constructions and results of this section easily extend to the face homology of topological pairs ( cf@xmath134 remark [ singsing ] ) .
the aim of this section is to reformulate face homology of path spaces in terms of smooth polychains . as a warm up
, we start with a study of smooth polychains in a smooth manifold .
recall from section [ manifolds with faces ] that a map @xmath1090 from an @xmath405-dimensional manifold with faces @xmath409 to a smooth @xmath435-dimensional manifold @xmath0 ( possibly , with boundary ) is said to be smooth if restricting @xmath73 to any local coordinate systems in @xmath409 and @xmath0 we obtain a map that extends to a @xmath407-map from an open subset of @xmath436 to @xmath437 .
if @xmath1091 is a union of ( some ) faces of @xmath409 , then we call a map @xmath1092 _ smooth _ whenever its restrictions to all faces of @xmath409 contained in @xmath1093 are smooth .
such a map @xmath1094 is necessarily continuous .
this terminology applies in particular to @xmath1095 .
[ smo1 ] let @xmath409 be a manifold with faces .
any smooth map @xmath1096 extends to a smooth map from a neighborhood of @xmath410 in @xmath409 to @xmath1097 . using a partition of unity on @xmath409 and local coordinates on @xmath0
, we easily reduce the lemma to the case where @xmath1098 with @xmath406 and @xmath1099 .
we need to prove that every function @xmath1100 whose restrictions to all proper faces of @xmath1101 are smooth extends to a smooth function on @xmath1101 .
we exhibit one such extension explicitly . for a subset @xmath982 of the set @xmath1102 and a point @xmath1103 denote by @xmath1104 the point of @xmath1101 whose
@xmath496-th coordinate is @xmath1105 if @xmath1106 and zero otherwise . if @xmath1107 , i.e@xmath134 if @xmath982 is a proper subset of @xmath1108 , then @xmath1109 . set @xmath1110 each function @xmath1111 is smooth because it is a composition of @xmath259 with the projection of @xmath1101 onto its proper face . therefore the function @xmath1112 is smooth .
moreover , it satisfies @xmath1113 for all @xmath1114 .
indeed , pick @xmath1115 such that @xmath1116 and observe that each term in corresponding to @xmath1117 with @xmath1118 cancels with the term corresponding to @xmath1119 provided the latter set is proper .
this leaves only the term @xmath1120 determined by @xmath1121 .
[ smo2 ] let @xmath1122 be a continuous map from a manifold with faces @xmath409 to a smooth manifold @xmath0 .
then any homotopy of @xmath1123 to a smooth map extends to a homotopy of @xmath73 to a smooth map @xmath1124 .
observe first that if @xmath1123 is smooth , then there is a homotopy of @xmath73 rel @xmath410 to a smooth map @xmath1125 . indeed
, lemma [ smo1 ] yields an extension of @xmath1123 to a smooth map @xmath1126 where @xmath1127 is a collar of @xmath410 in @xmath409 .
the latter map obviously extends to a continuous map @xmath1128 homotopic to @xmath73 rel @xmath410 .
since @xmath1129 is smooth and @xmath1130 is a compact subset of the smooth manifold @xmath1131 , there is a homotopy of @xmath1132 to a smooth map @xmath1133 , and this homotopy may be chosen to be constant in a neighborhood of @xmath410 in @xmath1127 .
the resulting smooth map @xmath1133 is homotopic to @xmath73 rel @xmath410 . to prove the lemma ,
take an arbitrary ( continuous ) extension of the given homotopy of @xmath1134 to a homotopy of @xmath1090 , and compose it with a homotopy @xmath1135 of the resulting map @xmath1136 to a smooth map as in the previous paragraph .
next , using a collar of @xmath410 in @xmath409 , deform the composed homotopy of @xmath73 into a homotopy satisfying the conditions of the lemma . as an exercise
, the reader may deduce from lemma [ smo2 ] ( by an inductive construction on the faces of @xmath409 ) that any continuous map @xmath1125 is homotopic to a smooth map .
a polychain @xmath895 in @xmath0 is _ smooth _ if the map @xmath1137 is smooth .
we explain now how to deform arbitrary polychains in @xmath0 into smooth polychains . for the notion of a deformation of a polychain in @xmath0 ,
see section [ homotopy_singular ] .
we explain first how to extend deformations .
let @xmath1138 consist of some faces of @xmath409 .
we say that a homotopy @xmath1139 of @xmath1140 is _ compatible _ with the partition @xmath443 if @xmath1141 for any @xmath689 and any faces of the same type @xmath1142 .
[ sssppsmodefo1- ] let @xmath634 be a polychain in @xmath0 and let @xmath1093 be a union of faces of @xmath409 .
let @xmath1139 be a homotopy of @xmath1143 compatible with @xmath443 such that @xmath1144 is smooth .
then there is a deformation @xmath684 of @xmath1145 such that @xmath1146 is smooth and @xmath1147 for all @xmath689 . for any integer @xmath418 , denote ( as in section [ manifolds with faces ] ) by @xmath1148 the union of all faces of @xmath409 of dimension @xmath476 .
recursively in @xmath1149 , we construct a homotopy @xmath1150 of @xmath1151 to a smooth map @xmath1152 . for @xmath1153
, there is nothing to do since @xmath1154 .
the induction step goes as follows . for each type of @xmath418-dimensional faces of @xmath409 , select a representative face @xmath439 so
that if at least one face of the given type lies in @xmath1093 , then @xmath1155 . if @xmath1155 , then set @xmath1156 for all @xmath1157 .
if @xmath1158 , then by lemma [ smo2 ] there is a homotopy of @xmath1159 to a smooth map extending the homotopy of @xmath73 on @xmath1160 obtained at the previous step .
the homotopy on the selected @xmath418-dimensional faces uniquely extends to a homotopy of @xmath73 on @xmath475 compatible with @xmath443 . for @xmath1161
, we obtain the required deformation of @xmath73 .
[ ppsmodefo1- ] for any polychain @xmath634 in @xmath0 , there is a deformation @xmath684 of @xmath1145 such that @xmath1146 is smooth and @xmath1162 for all @xmath689 and all faces @xmath439 of @xmath409 on which @xmath1163 is smooth .
this is a special case of lemma [ sssppsmodefo1- ] where @xmath1093 is the union of all faces of @xmath409 on which @xmath1163 is smooth and @xmath1164 is the constant homotopy .
we can now reformulate the face homology of @xmath0 in terms of smooth polychains . note that if a polychain @xmath580 in @xmath0 is smooth , then so are the polychains @xmath1165 , @xmath1166 , and @xmath606 .
disjoint unions of smooth polychains are smooth . applying the definitions of section [ facehomology ] to @xmath1167 but considering only smooth polycycles and smooth polychains we obtain _ smooth face homology _ @xmath1168 .
[ + + smoothdefo1- ] the natural linear map @xmath1169 is an isomorphism .
lemmas [ homotimplieshomol ] and [ ppsmodefo1- ] imply that any polycycle in @xmath0 is homologous to a smooth polycycle .
this proves the surjectivity of the map in the statement of the theorem . to prove the injectivity it suffices to show that , for any homologous reduced smooth @xmath405-polychains @xmath623 ,
@xmath1170 in @xmath0 there are smooth @xmath626-polychains @xmath1171 in @xmath0 such that @xmath1172 . by assumption , there are @xmath626-polychains @xmath1173 in @xmath0 and a diffeomorphism @xmath1174 . for @xmath1175 ,
set @xmath1176 and let @xmath1177 be the union of the components of @xmath1178 underlying @xmath1179 .
lemma [ ppsmodefo1- ] yields a homotopy @xmath1180 of @xmath1181 to a smooth map @xmath1182 in the class of maps compatible with the partition @xmath1183 , which is constant on all faces of @xmath1184 on which @xmath1185 is smooth .
in particular , this homotopy is constant on @xmath1186 .
consider the homotopy @xmath1187 of @xmath1188 .
this homotopy is compatible with the partition @xmath1189 and is constant on @xmath1190 ( because @xmath1191 is smooth on @xmath1192 ) . clearly , @xmath1193 is a diffeomorphism of the smooth polychains @xmath1194 and @xmath1195 . for @xmath1175 ,
the polychain @xmath1196 is obtained from @xmath1197 by a deformation which is constant on @xmath1179 and transforms @xmath1198 into a smooth polychain .
to finish the proof , we need only to show that this deformation of @xmath1198 extends to a deformation of @xmath1199 into a smooth polychain .
this is done in 3 steps .
first of all , applying lemma [ sssppsmodefo1- ] to @xmath1200 and taking for @xmath34 the union of all connected components of non - zero weight we obtain that our deformation of @xmath1201 extends to a deformation of @xmath1202 into a smooth polychain .
the latter deformation induces a deformation of @xmath1203 into a smooth polychain .
one more application of lemma [ sssppsmodefo1- ] allows us to extend the latter deformation to a deformation of @xmath1204 into a smooth polychain @xmath1205
. then @xmath1172 .
pick two points @xmath1206 in a smooth manifold @xmath0 ( possibly , @xmath1207 and @xmath1208 ) .
a _ path _ in @xmath0 from @xmath62 to @xmath1209 is a continuous map @xmath643\to m$ ] carrying @xmath1210 to @xmath60 and @xmath300 to @xmath63 .
let @xmath1211 be the space of such paths with compact - open topology .
note that a map @xmath838 from a topological space @xmath409 to @xmath1212 is continuous if and only if the adjoint map @xmath1213 , carrying any pair @xmath1214 to @xmath1215 is continuous ; see , for example , ( * ? ? ?
* section 1.2.7 ) .
a map from a manifold with faces @xmath409 to any subspace @xmath1216 of @xmath1217 is _ smooth _ if the adjoint map @xmath1218 is smooth in the sense of section [ smoothmaps ] .
a polychain @xmath1219 in @xmath81 is if @xmath1220 is smooth .
applying the definitions of section [ facehomology ] but considering only smooth polycycles and smooth polychains in @xmath81 , we obtain _ smooth face homology _ @xmath1221 . in the next theorem
, we consider @xmath1222 .
[ wwdefo1- ] the natural linear map @xmath1223 is an isomorphism .
we follow the lines of section [ smoothmaps ] with @xmath0 replaced by @xmath87 .
first , we show that given a manifold with faces @xmath409 , any smooth map @xmath1224 extends to a smooth map from a neighborhood of @xmath410 in @xmath409 to @xmath1225 .
indeed , the adjoint map @xmath1226 extends to a smooth map @xmath1227 by sending @xmath1228 to @xmath60 and @xmath1229 to @xmath63 . by lemma [ smo1 ]
, the latter map extends to a smooth map @xmath1230 for some neighborhood @xmath1127 of @xmath1231 in @xmath1232 . clearly , @xmath1233 for a neighborhood @xmath1234 of @xmath410 in @xmath409 .
the map @xmath1235 is adjoint to a smooth map @xmath1236 extending @xmath259 . lemmas [ smo2][ppsmodefo1- ] remain true with @xmath0 replaced by @xmath87 .
the proofs above apply with the only difference that lemma [ smo1 ] should be replaced by the result of the previous paragraph .
the proof of theorem [ + + smoothdefo1- ] also works with @xmath0 replaced by @xmath87 .
this gives the desired result . in the case where @xmath1237 , the path space @xmath1238 is homotopy equivalent to a smaller space .
let @xmath1239 be the subspace of @xmath1240 consisting of all paths @xmath1241 from @xmath60 to @xmath63 such that @xmath1242 .
[ loopsp ] the inclusion map @xmath1243 is a homotopy equivalence .
we begin with an observation in set - theoretic topology .
consider a topological pair @xmath1244 and suppose that there is a homotopy @xmath1245 of the identity map @xmath1246 such that @xmath1247 for all @xmath1248 .
then the inclusion @xmath1249 is a homotopy equivalence and its homotopy inverse @xmath1250 is obtained from @xmath1251 by reducing the image to @xmath683 .
indeed , the family @xmath1252 is a homotopy between @xmath1253 and @xmath1254 .
since @xmath1255 for all @xmath689 , we have the family of maps @xmath1256 .
this is a homotopy between @xmath1257 and @xmath1258 . using a tubular neighborhood of @xmath92 in @xmath0
, we can easily construct a ( smooth ) family of embeddings @xmath1259 such that @xmath1260 if @xmath1261 or @xmath1262 , and @xmath1263 for all other pairs @xmath1264 . given @xmath689 and a path @xmath1241 from @xmath60 to @xmath63 , we define a path @xmath1265 by @xmath1266 for all @xmath1267 .
this gives a family of paths @xmath1268 such that @xmath1269 and @xmath1270 for all @xmath1248 .
the formula @xmath1271 defines a homotopy @xmath1272 of @xmath1273 such that @xmath1274 for all @xmath1248 .
now , the result of the previous paragraph implies that the inclusion @xmath1243 is a homotopy equivalence .
lemma [ loopsp ] implies that @xmath1275 .
the following theorem computes the face homology of @xmath1276 and @xmath87 in terms of smooth polychains in @xmath1276 .
[ loopsp+++++ ] the natural linear map @xmath1277 is an isomorphism .
consider the homotopy @xmath1278 of @xmath1273 introduced in the proof of lemma [ loopsp ] and set @xmath1279 . for any manifold with faces @xmath337 and for any map @xmath1280 , the adjoint map @xmath1281 of @xmath1282 is given by @xmath1283 consequently , if @xmath74 is smooth , then @xmath1284 is smooth for all @xmath689 . we conclude that for any smooth polychain @xmath675 in @xmath87 , the polychain @xmath1285 in @xmath1276 is smooth . by the surjectivity of the map in theorem [ wwdefo1- ] , any polycycle @xmath580 in @xmath1276 is homologous in @xmath87 to a smooth polycycle @xmath1286 in @xmath87 .
applying @xmath259 , we obtain that @xmath1287 is homologous in @xmath1276 to the smooth polycycle @xmath1288 .
the homotopy @xmath1289 induces a deformation of @xmath580 into @xmath1287 in @xmath1276 .
therefore @xmath580 is homologous to @xmath1287 in @xmath1276 .
thus , @xmath580 is homologous to @xmath1288 in @xmath1276 .
this proves the surjectivity of the natural map @xmath1277 . to prove the injectivity ,
consider two reduced smooth @xmath405-polycycles @xmath623 , @xmath1170 in @xmath1276 that are homologous in @xmath1276 .
then they are homologous in @xmath86 . by the injectivity of the map in theorem [ wwdefo1- ] ,
there are smooth @xmath626-polychains @xmath627 in @xmath86 such that @xmath1290 . applying @xmath259
we obtain that @xmath1291 where @xmath1292 are smooth polychains in @xmath1276 .
hence @xmath1293 .
the homotopy @xmath1294 induces a smooth deformation of @xmath1179 into @xmath1295 and therefore @xmath1296 for @xmath1175 .
we deduce that @xmath1297 .
this completes the proof of the injectivity of the natural map @xmath1277 and of the theorem .
in this section we explain how to present pairs of face homology classes of path spaces by smooth polychains which are transversal in an appropriate sense .
fix a smooth manifold @xmath0 possibly with boundary .
the @xmath1298 is a smooth manifold canonically diffeomorphic to @xmath0 .
we say that a smooth map @xmath1299 from a manifold with faces @xmath34 to @xmath1300 is if we have @xmath1301 and the restriction of @xmath1299 to @xmath1302 is transversal to @xmath1303 in the usual sense of differential topology .
( the interiors of @xmath34 and @xmath1304 are smooth manifolds so that the latter condition makes sense . )
furthermore , @xmath1299 is said to be if its restriction to any face of @xmath34 is weakly transversal to the diagonal .
fix four points @xmath1305 which are distinct except that possibly @xmath1306 and/or @xmath1307 . in this subsection
we study transversality for smooth maps from manifolds with faces to the spaces @xmath1308 and @xmath1309 . fix manifolds with faces @xmath409 and @xmath337 .
consider smooth maps @xmath1310 , @xmath1311 and let @xmath1312 , @xmath1313 be the adjoint maps . the product map @xmath1314 carries a tuple @xmath1315 to @xmath1316 .
the latter point can lie on @xmath1317 only when @xmath1318 and never lies in @xmath1319 .
we say that @xmath73 and @xmath1320 are if the map @xmath1321 is transversal to @xmath1304 in the sense defined above .
note that the maps @xmath73 and @xmath1320 are transversal if and only if they are transversal in the sense of @xcite : see proposition 7.2.2 therein .
( this notion of transversality is stronger than the notion used in @xcite : see remark 6.3 therein . )
if @xmath1322 and @xmath1323 are transversal , then their restrictions to arbitrary faces of @xmath409 , @xmath337 are transversal .
clearly , @xmath1324-small smooth homotopies of @xmath73 and @xmath74 preserve transversality . here ,
for @xmath1325 , we say that a smooth homotopy of @xmath73 is if for any @xmath1326 ( and some riemannian metrics on @xmath0 and @xmath409 ) , the adjoint map @xmath1327 of @xmath73 and all its partial derivatives of length @xmath476 are changed during the homotopy by at most @xmath1328 at all points .
[ defo1 ] for any smooth maps @xmath1329 and @xmath1330 , there is an arbitrarily @xmath407-small smooth homotopy @xmath696 of @xmath687 such that @xmath1331 and @xmath74 are transversal .
proceeding by induction on @xmath1332 , we can assume that @xmath73 is transversal to the restrictions of @xmath74 to all proper faces of @xmath337 .
all subsequent homotopies of @xmath73 are chosen to be small enough to preserve this property .
fix a smooth triangulation @xmath469 of @xmath1333 .
a map from a simplex @xmath305 of @xmath469 to @xmath0 is if it is smooth as a singular simplex in @xmath0 .
a smooth map @xmath1334 is if the map @xmath1335 is weakly transversal to @xmath1336 .
we call a map @xmath1337 if the adjoint map from @xmath409 to the space of paths in @xmath0 takes values in @xmath1338 .
the map @xmath1299 is _
@xmath469-good _ if it is good and the image of any simplex of @xmath469 under @xmath1299 lies in a closed ball in @xmath0 .
consider the map @xmath1339 adjoint to @xmath1340 .
clearly , @xmath1341 is good .
since @xmath1342 , the set @xmath1343 is disjoint from @xmath1344 . by continuity , there is a small @xmath1345 such that the sets @xmath1346)$ ] and @xmath1347)$ ] are disjoint from @xmath1348 . subdividing @xmath469
, we can assume that @xmath1349 is @xmath469-good and @xmath1350\cup [ 1-\delta , 1])$ ] is a subcomplex of @xmath469 .
for any simplex @xmath305 of @xmath1351 , the map @xmath1352 is @xmath74-transversal because @xmath1353 .
set @xmath1354 .
we shall construct @xmath1355 homotopies @xmath1356 in the class of @xmath469-good maps @xmath1357 such that for all @xmath1358 and any simplex @xmath305 of @xmath1359 of dimension @xmath1360 , the map @xmath1361 is @xmath74-transversal .
then , the restriction of @xmath1362 to any simplex of @xmath1359 is @xmath74-transversal . for each face @xmath1363 of @xmath409 , the product @xmath1364 is a subcomplex of @xmath469 .
therefore , the restriction of @xmath1365 to @xmath1366 is weakly transversal to @xmath1304 . by the beginning of the proof ,
the same holds when @xmath337 is replaced with any of its proper faces .
this shows that the smooth map @xmath1367 determined by @xmath1362 is transversal to @xmath74 .
the homotopies @xmath1368 are constructed recursively . for @xmath1369 ,
the condition on @xmath1370 is void , and we can take @xmath1371 .
assume that we have required homotopies @xmath1372 for some @xmath1358 .
consider an @xmath1373-dimensional simplex @xmath1374 .
clearly , @xmath1375 . for @xmath1376 , let @xmath1377 denote the ( closed ) metric @xmath1328-neighborhood of @xmath1378 in @xmath1379 .
the inductive assumption implies that _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
@xmath1380 for a sufficiently small @xmath1326 , the restriction of the map @xmath1381 to @xmath1382 is weakly transversal to @xmath1304 . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ since @xmath1370 is good and @xmath1375 , we have @xmath1383 . since @xmath1370 is @xmath469-good , @xmath1384 for a closed ball @xmath1385 .
we can choose @xmath24 so that @xmath1386 .
we identify @xmath24 with the closed unit ball in euclidean space with center @xmath183 .
pick a small neighborhood @xmath1387 of @xmath1388 so that @xmath1389 for all @xmath1390 .
consider the smooth maps @xmath1391 where @xmath1392 carries any @xmath1393 to @xmath1394 .
it is obvious that the family of maps @xmath1395 is weakly transversal to @xmath1304 in the sense that the adjoint map @xmath1396 is weakly transversal to @xmath1304 . by the transversality theorem ( see ( * ? ? ?
* section 2.3 ) ) , we obtain that _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath1397 the set @xmath1398 is dense ( and open ) in @xmath982 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ( this argument is adapted from that of laudenbach ( * ? ? ?
* proof of lemma 2.6 ) . ) in our terminology , @xmath1392 is @xmath74-transversal for any @xmath1399 . for each @xmath1400 , we define a map @xmath1401 by @xmath1402 for all @xmath1393 where @xmath1403 is a smooth function carrying @xmath1404 to @xmath183 and @xmath1405 to @xmath300 .
then @xmath1406 on @xmath1404 and @xmath1407 on @xmath1408 .
we deduce from @xmath1380 and @xmath1397 that , for all sufficiently small @xmath1399 , the map @xmath1409 is @xmath74-transversal .
pick such an @xmath1005 and consider the obvious linear homotopy @xmath1410 constant on @xmath1411 . combining such homotopies corresponding to all @xmath1373-dimensional simplices @xmath1374 and extending by the constant homotopy on @xmath1351
we obtain a homotopy of @xmath1370 on the union of @xmath1351 with the @xmath1373-skeleton of @xmath469 .
the latter homotopy extends to a homotopy @xmath1412 in the class of good maps @xmath1413 . taking the vectors @xmath1005 in this construction small enough , we can always choose the homotopy @xmath1412 so that it proceeds in the class of @xmath469-good maps .
since this homotopy is constant on @xmath1414 and on the @xmath1415-th skeleton of @xmath469 , the @xmath74-transversality of @xmath1370 on the simplices of @xmath469 of dimension @xmath1416 acquired at the previous steps is preserved during the homotopy . by construction ,
@xmath1417 is @xmath74-transversal on all @xmath1373-dimensional simplices of @xmath469 .
[ defo1 + ] the homotopy in lemma [ defo1 ] may be chosen to be constant on the union of all faces @xmath1363 of @xmath409 such that @xmath1418 is transversal to @xmath74 .
the proof proceeds by induction on @xmath1419 .
if @xmath1420 , then we take the constant homotopy on all connected components of @xmath409 on which @xmath1090 is transversal to @xmath74 and we take the homotopy provided by lemma [ defo1 ] on all other components of @xmath409 .
let @xmath102 be a positive integer such that the lemma holds for all @xmath409 of dimension @xmath1421 .
we prove the lemma for an arbitrary @xmath102-dimensional manifold with faces @xmath409 .
as above , if @xmath73 is transversal to @xmath74 on some connected components of @xmath409 , then we take the constant homotopy on that components .
thus we can assume without loss of generality that @xmath73 may be transversal to @xmath74 only on proper faces of @xmath409 .
let @xmath1422 be the set of all faces @xmath1363 of @xmath409 such that @xmath1423 is transversal to @xmath74 . by the definition of transversality , if @xmath1424 , then all faces of @xmath1363 also belong to @xmath1422 .
set @xmath1425 and note that @xmath1426 by our assumption .
all homotopies of @xmath1427 in the following construction are arbitrarily @xmath407-small smooth homotopies constant on @xmath1428 .
we recursively construct @xmath1429 homotopies @xmath1430 such that the restriction of @xmath1370 to any face of @xmath1431 of dimension @xmath1432 is transversal to @xmath74 for all @xmath1433 .
assume that we already have homotopies @xmath1434 with the required properties where @xmath1435 .
consider an @xmath1373-dimensional face @xmath1363 of @xmath409 not belonging to @xmath1422 . by the assumptions on @xmath1436 , the restriction of @xmath1437 to any proper face of @xmath1363 is transversal to @xmath74 .
since @xmath1438 , the inductive assumption guarantees that there is an arbitrarily @xmath407-small , smooth , constant on @xmath1439 homotopy of @xmath1440 into a map @xmath1441 transversal to @xmath74 .
combining these homotopies over all @xmath1363 as above together with the constant homotopy on @xmath1428 and extending to a small smooth homotopy of @xmath1442 on the rest of @xmath409 , we obtain a homotopy @xmath1443 with the required properties .
next pick a collar @xmath1444 of @xmath1445 in @xmath409 .
set @xmath1446 .
then @xmath1234 is a manifold with faces and @xmath1447 . by the above
, @xmath1448 is a smooth map whose restriction to all proper faces of @xmath409 is transversal to @xmath74 .
therefore , choosing the collar @xmath1127 sufficiently narrow , we can ensure that the map @xmath1449 is transversal to @xmath74 . by lemma [ defo1 ] ,
there is an arbitrarily @xmath407-small homotopy @xmath1450 of @xmath1451 such that @xmath1452 is transversal to @xmath74 .
this homotopy extends to a small homotopy @xmath696 of @xmath1453 constant on a neighborhood of @xmath410 in @xmath1127 .
if the homotopy @xmath1450 is sufficiently small , then the extension may be chosen so that @xmath1454 is transversal to @xmath74 for all @xmath689 .
then @xmath1455 is transversal to @xmath74 , and so , @xmath1456 is transversal to @xmath74 .
the composite homotopy @xmath1457 satisfies all the conditions of the lemma .
let @xmath1458 and @xmath1459 be as in section [ transs ] .
we say that smooth polychains @xmath1219 in @xmath1460 and @xmath1461 in @xmath1462 are if the maps @xmath1463 and @xmath1464 are transversal .
the following two lemmas show that any smooth polychain can be made transversal to a given smooth polychain by a small deformation .
[ defo1++eee ] let @xmath634 and @xmath1465 be smooth polychains in @xmath1458 and @xmath1459 , respectively .
let @xmath1093 be a union of faces of @xmath409 .
let @xmath1466 be a smooth homotopy of @xmath1143 compatible with @xmath443 such that @xmath1467 is transversal to @xmath675 .
then there is a smooth deformation @xmath684 of @xmath1145 such that @xmath1146 is transversal to @xmath675 and @xmath1468 for all @xmath689 .
we apply the same recursive method as in the proof of lemma [ sssppsmodefo1- ] with @xmath0 replaced by @xmath1458 .
the homotopy of @xmath1090 on a representative face @xmath439 is obtained in two steps .
first , we take an arbitrary smooth homotopy of @xmath1159 extending the homotopy of @xmath1469 obtained at the previous step
. then we compose with an additional smooth homotopy rel @xmath1470 to a map @xmath1471 transversal to @xmath74 .
the latter homotopy is provided by lemma [ defo1 + ] .
[ defo1++ ] let @xmath634 and @xmath1465 be smooth polychains in @xmath1458 and @xmath1459 , respectively .
there exists an arbitrarily @xmath407-small smooth deformation @xmath684 of @xmath565 such that the polychain @xmath1146 is transversal to @xmath675 and @xmath1162 for all @xmath689 and all faces @xmath439 of @xmath409 on which @xmath73 is transversal to @xmath74 .
this is a special case of lemma [ defo1++eee ] where @xmath1093 is the union of all faces of @xmath409 on which @xmath73 is transversal to @xmath74 and @xmath1472 is the constant homotopy . that the deformation @xmath1473 may be chosen arbitrarily @xmath407-small follows from lemmas [ defo1 ] and [ defo1 + ] .
we say that a pair of face homology classes @xmath1474 is by a pair ( @xmath1475 ) if @xmath580 is a smooth reduced polycycle in @xmath1458 representing @xmath32 , @xmath675 is a smooth reduced polycycle in @xmath1459 representing @xmath122 , and @xmath580 is transversal to @xmath675 .
the following lemma will play a key role in the sequel .
[ keytheor ] every pair @xmath1474 can be transversely represented by a pair of polycycles .
any two pairs of polycycles transversely representing @xmath1476 can be related by a finite sequence of transformations @xmath1477 of the following types : * @xmath1478 and @xmath1479 or @xmath1480 where @xmath968 is a smooth polychain in @xmath1458 transversal to @xmath675 ; * @xmath1481 and @xmath1482 or @xmath1483 where @xmath1484 is a smooth polychain in @xmath1459 transversal to @xmath580 .
the first claim follows from the surjectivity in theorem [ loopsp+++++ ] and lemma [ defo1++ ] . that we need only reduced polycycles follows from the fact that the reduction of a ( smooth ) polycycle gives a homologous ( smooth ) polycycle .
we prove the second claim of the lemma .
consider pairs of reduced polycycles @xmath1485 and @xmath1486 transversely representing @xmath1487 .
since @xmath623 is homologous to @xmath1488 , we have @xmath1489 for some @xmath626-polychains @xmath1173 in @xmath1458 .
the injectivity in theorem [ loopsp+++++ ] ensures that @xmath1490 , @xmath1491 can be chosen to be smooth .
then there are smooth polychains @xmath1171 in @xmath1458 transversal to @xmath675 such that @xmath1492 .
these polychains are obtained from @xmath1490 , @xmath1491 using the same method as in the proof of theorem [ + + smoothdefo1- ] with the following modifications : @xmath1493 , smooth `` @xmath1494 transversal to @xmath675 '' , homotopy `` @xmath1495 smooth homotopy '' , lemma [ sssppsmodefo1- ] @xmath1495 lemma [ defo1++eee ] , lemma [ ppsmodefo1- ] @xmath1495 lemma [ defo1++ ] .
the move @xmath1496 expands as the composition of the following type ( i ) moves : @xmath1497 ( the middle move is a type ( i ) move corresponding to @xmath1498 . ) a similar argument shows that if two pairs of polycycles @xmath1499 and @xmath1500 transversely represent @xmath1487 , then the move @xmath1501 is a composition of type ( ii ) moves .
consider now any pairs of polycycles @xmath1502 and @xmath1503 transversely representing @xmath1487 . by lemma [ defo1++ ] ,
there is an arbitrarily @xmath407-small smooth deformation of @xmath1504 into a polycycle @xmath580 transversal to @xmath1505 .
we assume the deformation to be so small that @xmath565 is transversal to @xmath1506 as well .
by lemma [ homotimplieshomol ] , @xmath1507 represents @xmath32 . by the previous paragraph ,
each of the moves @xmath1508 expands as a composition of moves of types ( i ) and ( ii ) .
we define an intersection " operation for polychains in path spaces of a manifold . throughout this section
we use the following notation .
the symbol @xmath0 denotes an oriented smooth @xmath405-dimensional manifold with boundary , where @xmath1509 .
we fix four points @xmath1510 in @xmath92 .
we require that @xmath1511 but allow the equalities @xmath1306 and @xmath1307 . for any @xmath1512 , we set @xmath1513 and @xmath1514 .
let @xmath634 be a smooth polychain of dimension @xmath102 in @xmath1515 and let @xmath712 be a smooth polychain of dimension @xmath711 in @xmath1516 .
assume that @xmath580 and @xmath675 are transversal in the sense of section [ transversality_polychains ] .
we derive from @xmath580 and @xmath675 an intersection polychain " in @xmath1517 .
let @xmath1327 and @xmath1518 be the adjoint maps of @xmath1090 and @xmath1519 respectively . set @xmath1520 and consider the map @xmath1521 . since @xmath1522 and @xmath644 are manifolds with faces of dimensions @xmath233 and @xmath300 , respectively , @xmath34 is a manifold with faces of dimension @xmath1523 .
the transversality of @xmath73 and @xmath74 implies that @xmath1524 is empty if @xmath1525 and is a @xmath1526-dimensional manifold with corners if @xmath1527 . in the latter case each point of @xmath56 has a neighborhood @xmath1234 in @xmath34 such that @xmath1528 is homeomorphic to @xmath1529 for some integers @xmath1530 with @xmath1531 and @xmath1234 is homeomorphic to @xmath1532 .
these claims follow from the general theorems about transversality and about submanifolds of manifolds with corners , see ( * ? ? ?
* propositions 3.1.14 and 7.2.7 ) . consequently , @xmath1533 so that we can consider the commutative diagram of inclusion maps @xmath1534 ^ -{i } \ar[d]_-{j ' } & & \pi_0(p(n ) \cap v ) \ar[d]^-{j } \\ \pi_0(p(d ) ) \ar[rr]^-{i ' } & & \pi_0(p(n ) ) \ , . } \ ] ] the structure of @xmath338 described above implies that @xmath496 is a bijection between @xmath1535-element sets . since @xmath34 is a manifold with faces , @xmath810 is an injection
. therefore @xmath1536 is injective which implies that @xmath56 is a manifold with faces .
the faces of @xmath56 are the connected components of the intersections of @xmath56 with faces of @xmath34 .
we now upgrade @xmath56 to a polychain in @xmath1517 .
first of all , we orient @xmath56 as follows .
we use the orientation of @xmath1537 and the product orientation of @xmath1538 to orient the normal vector bundle of @xmath1317 in @xmath1538 ( see the introduction for our orientation conventions ) .
next , we pull - back this orientation of the normal vector bundle along @xmath1539 to obtain an orientation of the normal vector bundle of @xmath56 in @xmath34 .
the latter together with the product orientation in @xmath1540 induces an orientation of @xmath56 .
we also equip @xmath56 with the weight @xmath1541 which , for any connected components @xmath81 of @xmath409 and @xmath683 of @xmath337 , carries all connected components of @xmath56 contained in @xmath1542 to @xmath1543 .
next , we define a continuous map @xmath1544 by @xmath1545 for any @xmath1546 and @xmath1547 , where we set @xmath1548$}\\ 1 - 2 ( 1-\ell ) ( 1-u ) & \hbox{for $ \ell \in i , u\in [ 1/2,1]$. } \end{array}\right.\ ] ] the key property of the operation @xmath1549 is that for any @xmath1550 , we have @xmath1551 if @xmath1552 $ ] and @xmath1553 if @xmath1554 $ ] . for a fixed @xmath1555 , the point @xmath1556 moves along the path @xmath1557 from @xmath1558 to the diagonal point @xmath1559
as @xmath797 increases from @xmath183 to @xmath1560 and , next , it moves from that diagonal point to @xmath1561 along the path @xmath1562 as @xmath797 increases from @xmath1560 to @xmath300 : see figure [ diamond ] .
thus the map @xmath1563 is adjoint to a continuous map @xmath1564 whose coordinate maps are denoted by @xmath1565 and @xmath1566 .
[ t ] at 35 3 [ t ] at 466 3 [ t ] at 253 3 [ t ] at 143 6 [ t ] at 407 3 [ l ] at 542 8 at 32 116 [ r ] at 27 116 at 474 114 at 145 177 at 409 156 [ l ] at 479 113 at 253 236 at 28 436 [ r ] at 23 436 at 472 401 at 142 369 at 407 412 [ l ] at 478 401 [ r ] at 355 127 [ b ] at 155 415 [ r ] at 340 420 [ t ] at 194 140 for a fixed @xmath1555 and @xmath797 running from @xmath183 to @xmath300.,title="fig : " ] finally , we define a partition of @xmath56 . here
we need the assumption that the images of @xmath73 and @xmath74 lie in @xmath1567 and @xmath1568 , respectively .
this assumption implies that @xmath1569 .
therefore each face @xmath439 of @xmath56 is contained in a unique smallest face @xmath1570 of @xmath34 , where @xmath1571 is a face of @xmath409 and @xmath1572 is a face of @xmath337 .
note that the codimension of @xmath439 in @xmath56 is equal to the codimension of @xmath1573 in @xmath34 .
we declare two faces @xmath439 and @xmath451 of @xmath56 to have the same type if and only if @xmath1574 has the same type as @xmath1575 in @xmath409 , @xmath1576 has the same type as @xmath1577 in @xmath337 , and the diffeomorphism @xmath1578
^ -{\varphi_{a , a ' } \times { \operatorname{id}}_i \times \psi_{b , b ' } \times { \operatorname{id}}_i } & & & a ' \times i \times b ' \times i = n_{f ' } } \ ] ] carries @xmath439 onto @xmath451 . by restriction , we obtain a diffeomorphism @xmath1579 for any such @xmath439 , @xmath451 .
this defines a partition , @xmath1580 , of @xmath56 .
note that , for each face @xmath439 of @xmath56 , the faces of @xmath56 of the same type as @xmath439 are in one - to - one correspondence with the pairs @xmath1581 where @xmath887 is a face of @xmath409 of the same type as @xmath1571 and @xmath1582 is a face of @xmath337 of the same type as @xmath1572 .
the tuple @xmath1583 is a polychain in @xmath1584 .
we need only to show that the map @xmath1585 is compatible with the partition @xmath1580 .
consider two faces @xmath439 and @xmath451 of the same type in @xmath56 and set @xmath1574 , @xmath1576 , @xmath1575 and @xmath1577 .
for any @xmath1586 and @xmath1587 $ ] , we have @xmath1588 where the third equality follows from the compatibility of @xmath73 with @xmath443 and of @xmath74 with @xmath853 .
a similar argument works for @xmath1589 $ ] .
thus , @xmath1590 .
we study the behavior of the intersection operation @xmath1591 under the operations on polychains introduced in sections [ reducedpolychains ] and [ reducedpolychains ] .
[ reduction_d ] let @xmath1592 be smooth @xmath102-polychains in @xmath1593 and @xmath1594 be smooth @xmath711-polychains in @xmath1595 such that @xmath1592 and @xmath1594 are pairwise transversal
. then * @xmath1596 for any @xmath138 ; * @xmath1597 ; * @xmath1598 ; * @xmath1599 ; * @xmath1600 @xmath1601 . claims ( i ) and ( v ) are obvious . claim ( iv ) easily follows from ( ii ) and ( iii ) .
we prove ( ii ) .
it is clear that @xmath1602 and @xmath1603 using the identities @xmath1604 , we conclude that @xmath1605 we now prove ( iii ) .
let @xmath634 , @xmath712 and @xmath1606 as in section [ def_d ]
. consider the boundary polychain @xmath1607 as defined in section [ reducedpolychains ] , as well as the polychains @xmath1608 and @xmath1609 we verify that @xmath1610 consider a principal face @xmath439 of @xmath56 .
since the codimension of @xmath439 in @xmath56 is equal to the codimension of @xmath1611 in @xmath1612 , the face @xmath1613 is a principal face of @xmath34 .
therefore either @xmath1571 is a connected component of @xmath409 and @xmath1572 is a principal face of @xmath337 , or , @xmath1571 is a principal face of @xmath409 and @xmath1572 is a connected component of @xmath337 .
we first analyze the former case .
. then @xmath1615 corresponds to a connected component @xmath1616 of @xmath1617 via the map @xmath1618 where @xmath1619 is the natural map as in section [ reducedpolychains ] . the orientation of @xmath439 induced by that of @xmath1620 may differ from the orientation of @xmath1616 induced by @xmath1621 , and we now compute this difference .
let @xmath1622 be the trivial @xmath300-dimensional vector bundle equipped with the canonical orientation and let @xmath1623 be the same bundle with the opposite orientation .
given a cartesian product of topological spaces , let @xmath1624 denote the projection onto the @xmath496-th factor .
set @xmath1625 , which is a submanifold with faces of @xmath1626 of codimension @xmath183 containing @xmath1616 .
we can also view @xmath1627 as a submanifold of @xmath34 of codimension @xmath300 , so that @xmath1628 corresponds to @xmath1629 . using the orientation conventions of the introduction and using the letter @xmath469 for the tangent vector bundle of a manifold , we obtain the following orientation - preserving isomorphisms of oriented vector bundles : @xmath1630 therefore @xmath1631 where the letter @xmath1632 stands for the normal vector bundle of a submanifold in the ambient manifold . on the other hand , restricting the orientation - preserving isomorphism of oriented vector bundles @xmath1633 to @xmath439 we obtain that @xmath1634 since the orientations of @xmath1635 and @xmath1636 are both induced by the orientation of the normal bundle of @xmath1317 inx @xmath1538 , the bundle isomorphism @xmath1637 induced by the inclusion @xmath1638 is orientation - preserving . combining with the computations above , we deduce that @xmath1639 .
the case where @xmath1571 is a principal face of @xmath409 and @xmath1572 is a connected component of @xmath337 is treated similarly . in this case
, @xmath439 corresponds to a connected component @xmath1640 of @xmath1641 , and the orientation of @xmath1642 induced from that of @xmath56 differs by @xmath1643 from the orientation of @xmath1644 induced by @xmath1641 .
this gives the diffeomorphism of oriented manifolds with faces , which is easily checked to be a diffeomorphism of polychains as in ( iii ) .
we introduce an operation @xmath1645 in the face homology of path spaces of manifolds .
we keep notation of section [ notat ] .
first we show that the intersection operation @xmath1591 of section [ operation_d ] induces an operation on the level of face homology .
[ upsilon_tilde_def ] for any integers @xmath1646 , the intersection @xmath1647 defined in section [ def_d ] induces a bilinear map @xmath1648 consider any face homology classes @xmath1649 and @xmath1650 . by lemmas [ loopsp ] and [ keytheor ] , the pair @xmath1651 can be transversely represented by a smooth reduced @xmath102-polycycle @xmath580 in @xmath1652 and a smooth reduced @xmath711-polycycle @xmath675 in @xmath1653 .
it follows from lemma [ reduction_d].(iv ) that the polychain @xmath1654 in @xmath1584 is a polycycle .
consider another such pair @xmath1655 transversely representing @xmath1651 .
we claim that the polycycles @xmath1654 and @xmath1656 are homologous . by lemma [ keytheor ] , it suffices to prove this claim in the following two cases : * @xmath1657 and there exists a smooth @xmath1658-polychain @xmath968 in @xmath1593 transversal to @xmath675 such that @xmath1659 or @xmath1660 ; * @xmath1661 and there exists a smooth @xmath1662-polychain @xmath1484 in @xmath1595 transversal to @xmath580 such that @xmath1663 or @xmath1664 .
assume for concreteness that @xmath1657 and @xmath1659 ( the other cases can be treated similarly ) .
since @xmath675 is a reduced polycycle , lemma [ reduction_d].(iv ) implies that @xmath1665 this and lemma [ reduction_d].(v ) imply that @xmath1666 we conclude that @xmath1667 is homologous to @xmath1668 .
thus , the face homology class @xmath1669 of @xmath1667 depends only on @xmath1670 and @xmath1671 .
this defines the pairing in the statement of the lemma .
the bilinearity of this pairing follows from assertions ( i ) and ( v ) in lemma [ reduction_d ] .
the pairing produced by lemma [ upsilon_tilde_def ] induces a linear map @xmath1672 taking the direct sum over all @xmath1673 , we obtain a linear map of degree @xmath1674 @xmath1675 to stress the role of the tuple of base points @xmath1676 we will also denote this map by @xmath1677 . any permutation @xmath1678 of @xmath1676 such that @xmath1679 yields a map @xmath1680 we now establish the following symmetry for @xmath1645 .
[ antisymmetry_upsilon_tilde ] let @xmath1681 be the map permuting the two factors of the cartesian product .
for any @xmath1649 and @xmath1682 with @xmath1673 , @xmath1683 we assume that @xmath1651 is transversely represented by a smooth reduced @xmath102-polycycle @xmath634 in @xmath1652 and a smooth reduced @xmath711-polycycle @xmath712 in @xmath1653 .
let @xmath1606 and @xmath1684 .
let @xmath1685 be the permutation map @xmath1686 .
this map preserves @xmath1317 pointwise and preserves ( respectively , inverts ) the orientation of the normal bundle of @xmath1317 in @xmath1538 if @xmath405 is even ( respectively , odd ) .
let @xmath1687 be the permutation map defined by @xmath1688 .
clearly , @xmath1689 and @xmath1690 .
thus , @xmath1691 restricts to a diffeomorphism @xmath1692 of degree @xmath1693 .
this diffeomorphism carries the weight @xmath1694 to @xmath1695 and the partition @xmath1580 to @xmath1696 . also ,
thus , @xmath1698 is a diffeomorphism of the polychains @xmath1699 and @xmath1700 .
we conclude that @xmath1701 to evaluate @xmath1702 on a pair of face homology classes in @xmath1703 , we represent these classes by smooth reduced transversal polycycles in @xmath1704 and take the face homology class of the intersection polycycle .
we now explain how to compute @xmath1702 from more general polycycles in @xmath1703 .
we say that polycycles ( possibly non - smooth and non - reduced ) @xmath1705 in @xmath1706 and @xmath712 in @xmath1707 are _ admissible _ if there exist open sets @xmath1708 and @xmath1709 such that * the maps @xmath1710 and @xmath1711 are smooth and their images do not meet @xmath92 ; * @xmath1712 ; * for any face @xmath1363 of @xmath409 and any face @xmath439 of @xmath337 , the restriction of @xmath1713 to @xmath1714 is transversal to @xmath1303 in the usual sense of differential topology .
if @xmath580 and @xmath675 are admissible , then we can define the intersection polycycle @xmath1715 in @xmath1716 repeating word for word the definitions of section [ operation_d ] .
the intersection polycycle depends only on @xmath580 , @xmath675 and does not depend on the choice of the sets @xmath1127 and @xmath1234 as above .
[ newcomputation ] let @xmath580 and @xmath675 be admissible polycycles in @xmath1706 and @xmath1707 representing , respectively , @xmath1717 and @xmath1718
. then @xmath1719 .
let @xmath1705 and @xmath712 .
the set @xmath1720 is closed and therefore compact . by ( ii ) , there are compact sets @xmath1721 and @xmath1722 such that @xmath1723 . pick a small deformation of @xmath73 and @xmath74 into smooth maps ( in the class of maps compatible with the partitions ) .
the deformation may be chosen to be constant on some open neighborhoods @xmath1724 , @xmath1725 of @xmath1726 , respectively , and to be so small that the condition ( ii ) with @xmath1727 replaced by @xmath1728 is met during the deformation . the condition ( iii ) with @xmath1729 replaced by @xmath1730 is automatically met during the deformation . by lemma [ homotimplieshomol ] ,
the face homology class @xmath1731 is preserved under such a deformation .
thus , without loss of generality we can assume from the very beginning that the maps @xmath73 and @xmath74 are smooth . pick a small neighborhood @xmath93 of @xmath92 in @xmath0 such that @xmath1732 .
the proof of lemma [ loopsp ] and theorem [ loopsp+++++ ] provides , for any @xmath1512 , a homotopy of the identity map @xmath1733 into a map @xmath1734 such that smooth polycycles in @xmath1735 remain smooth throughout the homotopy .
the homotopy acts on a path in @xmath0 from @xmath1736 to @xmath1737 by pushing the interior points of the path inside @xmath0 along a 1-parameter family of embeddings @xmath1738 .
we can assume that these embeddings are constant on @xmath1739 and so , the homotopy fixes all points of the paths lying in @xmath1740 . for @xmath1741 and @xmath1742 ,
these homotopies induce a smooth deformation of polycycles @xmath1743 , @xmath1744 where @xmath687 , @xmath1745 , @xmath1746 , @xmath1747 .
our assumptions ensure that @xmath1748 and @xmath1749 for all @xmath689 .
thus the set @xmath1750 does not depend on @xmath1157 , and @xmath1751 are admissible for all @xmath689
. then the polycycle @xmath1752 is obtained from the polycycle @xmath1753 by deformation .
hence , by lemma [ homotimplieshomol ] , @xmath1754 .
the polycycles @xmath1755 in @xmath1652 and @xmath1756 in @xmath1653 transversely represent the pair @xmath1757 .
we conclude that @xmath1758 where the third equality is given by lemma [ reduction_d].(ii ) .
we formulate for @xmath1645 a leibniz - type rule in the second variable .
( since @xmath1645 is symmetric in the sense of lemma [ antisymmetry_upsilon_tilde ] , a leibniz - type rule in the first variable easily follows . )
pick a fifth base point @xmath1759 . for any @xmath1760 ,
the concatenation of paths @xmath1761 induces a bilinear _ concatenation pairing _
@xmath1762 similarly , for any @xmath1763 , the map @xmath1764 induces bilinear pairings @xmath1765 @xmath1766 [ leibniz_upsilon_tilde ] if @xmath1767 , then for any @xmath1768 , @xmath1769 , and @xmath1770 with @xmath1771 , @xmath1772 let @xmath1773 be smooth polycycles in @xmath1774 which represent @xmath293 respectively . applying lemma [ defo1++ ] twice ( and choosing homotopy there sufficiently small ) , we can assume that @xmath580 is transversal to both @xmath675 and @xmath955 .
then @xmath1775 is represented by the following polycycle in @xmath1776 : @xmath1777 where @xmath1778 and the adjoint map @xmath1779 is computed by @xmath1780,\\ \widetilde\rho(r,2t-1 ) & \hbox{for } l\in l , r\in r , t \in [ 1/2,1 ] .
\end{array}\right.\ ] ] the polycycles @xmath580 and @xmath1484 are admissible in the sense of section [ computationoftildeupsilon ] : we can take @xmath1781 and @xmath1782 .
it follows from lemma [ newcomputation ] that @xmath1783 .
thus , to prove the lemma , it is enough to show that @xmath1784 to this end , we compare @xmath1785 with @xmath1786 consider the embedding @xmath1787 defined by @xmath1788 and the embedding @xmath1789 defined by @xmath1790 . note that @xmath1791 has degree @xmath1792 while @xmath1793 has degree @xmath1794 .
consider also the cartesian projections @xmath1795 @xmath1796 clearly , @xmath1797 .
therefore , the map @xmath1791 restricts to a diffeomorphism @xmath1798 of degree @xmath1792 .
similarly , since @xmath1799 , the map @xmath1793 restricts to a diffeomorphism @xmath1800 of degree @xmath1801 .
here we use the following general fact involving our orientation conventions stated in the introduction : if @xmath81 , @xmath683 are oriented manifolds and @xmath982 is an oriented submanifold of @xmath81 , then the bundle map @xmath1802 induced by the cartesian projection @xmath1803 is an orientation - preserving isomorphism on each fiber , while the bundle map @xmath1804 induced by the cartesian projection @xmath1805 is orientation - preserving if and only if the product @xmath1806 is even .
it is clear from the definition of @xmath1484 and the computations of degrees above that @xmath1807 is an orientation - preserving diffeomorphism .
we claim that it transports the polychain structures of @xmath1808 and @xmath1809 into the polychain structure of @xmath1810 up to deformation of the latter .
this will imply and the lemma . to prove our claim , we need to verify that @xmath1811 preserves the face partitions and the weights and commutes with the maps to the path spaces up to deformation .
we start with the face partitions .
let @xmath1812 be faces of @xmath1813 of the same type and let @xmath1814 be faces of @xmath519 of the same type .
then @xmath1815 and @xmath1816 are faces of @xmath1817 of the same type .
we claim that the faces @xmath1818 and @xmath1819 of @xmath56 have the same type . by section [ def_d ]
, @xmath1820 is the restriction of the diffeomorphism @xmath1821 to @xmath451 where @xmath1822 ( respectively , @xmath1823 ) is the smallest face of @xmath1824 containing @xmath451 ( respectively , @xmath452 ) .
the smallest faces @xmath1573 and @xmath1825 of @xmath1826 containing @xmath439 and @xmath12 respectively are @xmath1827 clearly , the diagram @xmath1828_{p ' } \ar[rrrrr]^{\left(\varphi_{a_{f'},a_{g ' } } \times { \operatorname{id}}\times \psi_{b_{f'},b_{g ' } } \times { \operatorname{id}}\right ) \times \chi_{h , j } } & & & & & n_{g ' } \times j \ar[d]^{p ' } \\ n_f \ar[rrrrr]^{\varphi_{a_{f'},a_{g ' } } \times { \operatorname{id}}\times ( \psi_{b_{f'},b_{g'}}\times \chi_{h , j } ) \times { \operatorname{id } } } & & & & & n_g } \ ] ] commutes , so that the bottom diffeomorphism in that diagram carries @xmath439 onto @xmath12 .
we deduce that @xmath439 and @xmath12 have the same type in @xmath56 and the identification map @xmath1829 ( which , by definition , is the restriction of the bottom diffeomorphism to @xmath439 ) satisfies @xmath1830 this proves that @xmath1791 carries the partition @xmath1831 on @xmath1817 to the partition @xmath1580 restricted to @xmath1832 .
a similar argument shows that @xmath1793 carries the partition @xmath1833 on @xmath1834 to the partition @xmath1580 restricted to @xmath1835 .
it remains only to observe that a face of @xmath56 lying in @xmath1836 can not have the same type as a face of @xmath56 lying in @xmath1837 . to see this , we use the fact that every face @xmath439 of @xmath56 determines a smallest face @xmath1838 of @xmath1839 such that @xmath1840 and @xmath1841 are faces of @xmath1842 respectively .
if @xmath449 are faces of @xmath56 of the same type , then @xmath1841 must have the same type as @xmath1843 respectively , and the diffeomorphism @xmath1844 carries @xmath439 onto @xmath12 . since this diffeomorphism preserves the last coordinate and @xmath1845 , \quad ' \!p(l\times\ , ' \!d ) \subset k \times i \times ( l\times r ) \times [ 1/2,1 ] \ ] ] we deduce that @xmath439 and @xmath12 are both contained either in @xmath1836 or in @xmath1837 .
we next show that the diffeomorphism @xmath1811 preserves the weights .
let @xmath1846 be a connected component of @xmath1813 and let @xmath1057 be a connected component of @xmath519 .
the weight of the connected component @xmath1847 of @xmath1848 is @xmath1849 where @xmath1127 and @xmath1234 are connected components of @xmath409 and @xmath337 , respectively , such that @xmath1850 .
clearly , @xmath1851 so that @xmath1852 this proves that @xmath1791 carries the weight @xmath1853 on @xmath1817 to the weight @xmath1694 restricted to @xmath1854 .
a similar argument shows that @xmath1793 carries the weight @xmath1855 on @xmath1856 to the weight @xmath1694 restricted to @xmath1857 .
we now show that @xmath1811 commutes with the maps to @xmath1858 up to deformation .
the maps in question are @xmath1859 and @xmath1860 where @xmath1861 @xmath1862 we first compute @xmath1863 .
pick any @xmath1864 and @xmath1865 . for @xmath1866 $ ] , @xmath1867 similarly , for @xmath1868 $ ] , @xmath1869 we now separate two cases depending on whether or not @xmath1870 or , equivalently , on whether or not @xmath1871 . for @xmath1872
$ ] , we obtain @xmath1873 for @xmath1874 $ ] , we obtain @xmath1875 these computations show that the first coordinate map @xmath1876 of @xmath1877 is equal to @xmath1878 , which is also the first coordinate of the map @xmath259 given by .
the second coordinate maps @xmath1879 of @xmath1877 and @xmath259 may differ .
nonetheless , they are homotopic in the following way . for any @xmath1880 , consider the numbers @xmath1881 and let @xmath1882 \longrightarrow [ 0,s ] , \quad \beta_{t,{y } } : [ a_{y } , b_{t,{y } } ] \longrightarrow [ t,1 ] , \quad \gamma_{t,{y}}:[b_{t,{y}},1 ] \longrightarrow [ 0,1]\ ] ] be the affine maps carrying the left / right endpoints of segments to the left / right endpoints respectively .
we define a continuous map @xmath1883 by @xmath1884,\\ \widetilde \lambda(l,\beta_{t,{y}}({x } ) ) & \hbox{if } { x}\in [ a_{y},b_{t,{y}}],\\ \widetilde \rho(r,\gamma_{t,{y}}({x } ) ) & \hbox{if } { x } \in [ b_{t,{y}},1 ] .
\end{array}\right.\ ] ] observing that @xmath1885 and @xmath1886 , we conclude that @xmath1887 determines a homotopy between the second coordinate maps of @xmath259 and @xmath1877 in the class of maps @xmath1879 .
it remains to check that this homotopy is compatible with the partition @xmath1888 on @xmath1848 .
any faces @xmath449 of @xmath1848 of the same type expand as @xmath1889 and @xmath1890 where @xmath1812 are faces of @xmath1813 of the same type and @xmath1891 are faces of @xmath519 of the same type .
let @xmath1892 be the smallest faces of @xmath1824 containing @xmath451 and @xmath452 respectively .
the identifying map @xmath1893 is the restriction of the diffeomorphism @xmath1894 since the maps @xmath1895 are compatible with the partitions @xmath1896 respectively , we deduce from that for any @xmath1897 , @xmath1898 , and @xmath1899 , @xmath1900 hence for each @xmath1901 , the map @xmath1902 is compatible with the partition @xmath1831 .
we conclude that the homotopy of @xmath259 to @xmath1877 determined by @xmath305 is compatible with the partition @xmath1888 .
one similarly constructs a deformation of the map into @xmath1903 compatible with the partition .
consider another tuple @xmath1904 of points of @xmath92 such that @xmath1905 and set @xmath1906 .
section [ dfp_tildeupsilon ] yields a linear map @xmath1907 we compare @xmath1908 to the map @xmath1909 assuming that @xmath1736 and @xmath1910 belong to the same connected component of @xmath92 for all @xmath1911 .
choose a path @xmath1912 from @xmath1736 to @xmath1913 for each @xmath496 .
the formula @xmath1914 defines a continuous map @xmath1915 from @xmath1735 to @xmath1916 .
homotopic paths yield homotopic maps , and constant paths yield maps homotopic to the identity .
therefore @xmath1915 is a homotopy equivalence with homotopy inverse @xmath1917 .
the homotopy equivalence @xmath1915 induces an isomorphism in the face homology which we denote by the same symbol : @xmath1918 similarly , the isomorphism @xmath1919 induced by the homotopy equivalence @xmath1920 is also denoted by @xmath1920 .
[ naturbracket_tilde ] if @xmath1921 , then the following diagram commutes : @xmath1922 ^ - { \widetilde \upsilon } \ar[d]_-{(\varsigma_1,\varsigma_2)_\ # \otimes ( \varsigma_3 , \varsigma_4 ) _
\#}^-\simeq & \widetilde h_{\ast } ( \omega_{32 } \times \omega_{14 } ) \ar[d]^-{(\varsigma_3,\varsigma_2)_\ # \times ( \varsigma_1 , \varsigma_4 ) _ \#}_-\simeq \\ \widetilde h_\ast ( \omega'_{12 } ) \otimes \widetilde h_\ast ( \omega'_{34 } ) \ar[r]^-{\widetilde \upsilon ' } & \widetilde h_{\ast } ( \omega'_{32 } \times \omega'_{14 } ) . } \ ] ] since the isomorphism @xmath1923 depends only on the homotopy classes of the paths @xmath1924 , @xmath1925 , and since composition of the paths leads to composition of the corresponding isomorphisms , it is enough to consider the case where three of the paths @xmath1924 s are constant . assume for concreteness that @xmath1926 , @xmath1927 , @xmath1928 , and @xmath1929 are constant paths .
the assumption @xmath1921 implies that deforming if necessary the path @xmath1930 , we can ensure that @xmath1931 .
let @xmath1649 and @xmath1650 .
consider smooth polycycles @xmath1705 in @xmath1652 and @xmath1932 in @xmath1653 transversely representing the pair @xmath1933 .
the class @xmath1934 is represented by the polycycle @xmath1935 in @xmath1936 ( but not in @xmath1937 ) .
the polycycles @xmath580 and @xmath1938 are admissible in the sense of section [ computationoftildeupsilon ] : we can take @xmath1781 and @xmath1939
. set @xmath1940 and @xmath1941 where @xmath1942 .
it is easy to construct a diffeomorphism @xmath1943 preserving the orientation , the weight , and the face partition , and such that @xmath1944 is homotopic to @xmath1945 in the class of maps @xmath1946 compatible with @xmath1580 .
we deduce from lemma [ newcomputation ] that @xmath1947 this proves the commutativity of the diagram .
assuming that @xmath1921 , we extend the definition of @xmath1645 to arbitrary 4-tuples of points @xmath1948 , @xmath1949 , @xmath1950 , @xmath1951 . deforming these points in @xmath92
, we can obtain points @xmath1952 such that @xmath1953 . for @xmath1954 , pick a path @xmath1912 from @xmath1736 to @xmath1913 .
section [ dfp_tildeupsilon ] yields a linear map @xmath1955 where @xmath1906 for all @xmath136 .
then we define @xmath1956 to be the unique linear map such that the diagram commutes .
lemma [ naturbracket_tilde ] implies that this map depends neither on the choice of the paths @xmath1957 nor on the choice of the points @xmath1958 .
if @xmath1959 , then we can take @xmath1960 and the constant path @xmath1961 for all @xmath496 , and recover the same map @xmath1645 as before .
the properties of @xmath1645 established under the assumption @xmath1959 remain true for arbitrary base points in @xmath92 .
this easily follows from the definitions and the fact that the concatenation pairing is preserved under the change - of - base - points isomorphism . to conclude our study of the operation @xmath1645
, we define a renormalized version @xmath1962 by setting @xmath1963 for any homogeneous @xmath1964 and @xmath1965 .
the properties established above for @xmath1645 can be rephrased with the appropriate signs for @xmath1966 . in particular ,
lemma [ antisymmetry_upsilon_tilde ] yields the identity @xmath1967 where @xmath1968 is the @xmath405-degree .
besides , for any @xmath1969 ( distinct from @xmath1948 and @xmath1949 if @xmath1970 ) and any homogeneous @xmath1964 , @xmath1965 , @xmath1971 , lemma [ leibniz_upsilon_tilde ] yields the leibniz rule @xmath1972 finally , the diagram remains commutative with @xmath1645 replaced by @xmath1973 .
we derive from the map @xmath1973 introduced in section [ upsilon - tilde ] an operation @xmath1975 in singular homology of path spaces .
we keep notation of section [ notat ] but drop the condition @xmath1976 when @xmath1921 .
consider the linear map @xmath1977 defined by the commutative diagram @xmath1978 ^ - { \check\upsilon } & & \widetilde h_{\ast } ( \omega_{32 } \times \omega_{14 } ) \ar[d]^-{[-]}\\ h_\ast ( \omega_{12 } ) \otimes h_\ast ( \omega_{34 } ) \ar[u]^-{{\left\langle}-{\right\rangle}\times { \left\langle}-{\right\rangle } } \ar@{-->}[rr]^- { \upsilon } & & h_{\ast } ( \omega_{32 } \times \omega_{14 } ) . }
\ ] ] formula and the naturality of the transformation @xmath1979 : \widetilde h_\ast \to h_\ast $ ] imply the following antisymmetry of @xmath1975 : for any homogeneous @xmath1980 , @xmath1981 , @xmath1982 where @xmath1983 is the linear map induced by the permutation map @xmath1984 . if @xmath1921 , then the diagram with @xmath1645 replaced by @xmath1973 and the naturality of the transformations @xmath910 and @xmath1985 $ ] imply that the following diagram commutes : @xmath1986 ^ - { \upsilon } \ar[d]_-{(\varsigma_1,\varsigma_2)_\ # \otimes ( \varsigma_3 , \varsigma_4 ) _ \#}^-\simeq & h_{\ast } ( \omega_{32 } \times \omega_{14 } ) \ar[d]^-{(\varsigma_3,\varsigma_2)_\ # \times ( \varsigma_1 , \varsigma_4 )
_ \#}_-\simeq \\
h_\ast ( \omega'_{12 } ) \otimes h_\ast ( \omega'_{34 } ) \ar[r]^- { \upsilon ' } & h_{\ast } ( \omega'_{32 } \times \omega'_{14 } ) . } \ ] ] here , for every @xmath1987 , @xmath1988 is a point of @xmath92 connected to @xmath1736 by a path @xmath1989 , @xmath1990 is the map determined by the base points @xmath1991 , and , for all @xmath1992 , @xmath1915 stands for the homotopy equivalence @xmath1993 and for the induced isomorphism in singular homology . finally , the following crucial lemma will be proved in section [ subs123 ] .
[ diag - imp ] the following diagram commutes : @xmath1994_-{[- ] \times [ - ] } \ar[rr]^-{\check \upsilon } & & \widetilde h_\ast ( \omega_{32 } \times \omega_{14 } ) \ar[d]^-{[-]}\\ h_\ast ( \omega_{12 } ) \times h_\ast ( \omega_{34 } ) \ar[rr]^-{\upsilon } & & h_\ast ( \omega_{32 } \times \omega_{14 } ) . } \ ] ] as in section [ leibniz ] for face homology , the concatenation of paths induces three kinds of bilinear pairings for singular homology : @xmath1995 @xmath1996 @xmath1997 [ leibniz_upsilon ] for any @xmath1969 ( distinct from @xmath1948 and @xmath1949 if @xmath1970 ) and any homogeneous @xmath1998 , @xmath1999 , @xmath2000 , @xmath2001 for any @xmath2002 and @xmath2003 with @xmath2004 , we have @xmath2005 = [ { \mathsf{c}}_*(x \times y)]= { \mathsf{c}}_*[x\times y ] = { \mathsf{c}}_*([x]\times [ y])= [ x ] [ y]\ ] ] where we use the naturality of @xmath774 $ ] and lemma [ twocrossproducts ] . in particular , @xmath2006 [ { \left\langle}c { \right\rangle } ] = [ { \left\langle}b { \right\rangle}{\left\langle}c { \right\rangle}]$ ] .
we deduce that @xmath2007 \otimes \left [ { \left\langle}b{\right\rangle}{\left\langle}c{\right\rangle}\right]\big)\\ & = & \left[\check \upsilon_{12,35}\big({\left\langle}a { \right\rangle } , { \left\langle}b{\right\rangle}{\left\langle}c { \right\rangle}\big)\right]\\ & = & \left[\check \upsilon_{12,35}\big({\left\langle}a { \right\rangle } , { \left\langle}b{\right\rangle}\big ) { \left\langle}c { \right\rangle}+ ( -1)^{\vert a\vert_n \vert b\vert } { \left\langle}b{\right\rangle}\check \upsilon_{12,35}\big({\left\langle}a { \right\rangle } , { \left\langle}c { \right\rangle}\big ) \right ] \\ & = & \left[\check \upsilon_{12,35}\big({\left\langle}a { \right\rangle } , { \left\langle}b{\right\rangle}\big ) \right ]
c + ( -1)^{\vert a\vert_n \vert b\vert } b \left [ \check \upsilon_{12,35}\big({\left\langle}a { \right\rangle } , { \left\langle}c { \right\rangle}\big ) \right ] \\ & = & \upsilon_{12,35}(a , b ) c + ( -1)^{\vert a\vert_n \vert b\vert } b \upsilon_{12,35}(a , c)\end{aligned}\ ] ] where the second , third , fourth and fifth formulas follow respectively from , , , and the definition of @xmath1975 .
we claim that , for all @xmath1964 and @xmath2008 , @xmath2009{\right\rangle},b\big)\big ] = \big[\widetilde\upsilon(a , b)\big ] = \big[\widetilde\upsilon\big(a,{\left\langle}[b]{\right\rangle}\big)\big ] .\ ] ] this would imply similar equalities with @xmath2010 replaced by @xmath1973 .
therefore @xmath2011,[b ] ) = \left[\check \upsilon\big({\left\langle}[a]{\right\rangle},{\left\langle}[b ] { \right\rangle}\big)\right ] = \left[\check \upsilon\big(a,{\left\langle}[b ] { \right\rangle}\big)\right ] = \left[\check \upsilon(a , b)\right].\ ] ] this proves the commutativity of the diagram .
we prove the left equality in ; it would imply the right equality by the symmetry property of @xmath2010 ( lemma [ antisymmetry_upsilon_tilde ] ) . as explained in section [ checkupsilon ]
, we can assume that @xmath2012 .
we need to prove that , for any smooth polycycle @xmath634 in @xmath1652 and any smooth polycycle @xmath712 in @xmath1653 transversal to @xmath580 , @xmath2013 { \right\rangle},{\left\langle}{\mathscr{l}}{\right\rangle}\big)\big]=[{\mathscr{d}}({\mathscr{k}},{\mathscr{l } } ) ] .\ ] ] set @xmath2014 .
pick a locally ordered smooth triangulation @xmath469 of @xmath409 which fits @xmath443 .
the construction of such a triangulation in section [ fundcla ] ( using lemma [ partition ] ) shows that we can further assume that ( * ) @xmath2015 is a face of @xmath486 for any face @xmath439 of @xmath409 and any simplex @xmath486 of @xmath469 ( cf@xmath134 the last paragraph in the proof of lemma [ partition ] ) .
consider the fundamental @xmath102-chain @xmath2016 determined by @xmath469 as in section [ fundcla ] ( here @xmath534 runs over all @xmath102-simplices of @xmath469 ) .
then @xmath2017 is a smooth singular @xmath102-cycle in @xmath2018 representing the singular homology class @xmath826 \in h_p(\omega_{12}^\circ)$ ] .
next consider the smooth @xmath102-polycycle @xmath2019 in @xmath1652 associated with the expansion @xmath2020 as in section [ firsttransf ] . by construction
, @xmath560 is a disjoint union of copies of the standard @xmath102-simplex @xmath1063 indexed by @xmath102-simplices @xmath534 of @xmath469 and @xmath2021 where @xmath2022 . by the definition of the transformation @xmath2023 , @xmath2024 { \right\rangle}= { \left\langle}[\kappa_*(\sigma)]{\right\rangle}= { \left\langle}{\mathscr{k } } ' { \right\rangle}\in \widetilde h_{p}(\omega_{12}),\ ] ]
so that is equivalent to @xmath2025=[{\mathscr{d}}({\mathscr{k}},{\mathscr{l } } ) ] .\ ] ] lemma [ defo1++ ] yields a deformation of @xmath2026 into a polycycle @xmath2027 transversal to @xmath1286 .
such a polycycle @xmath2027 is also transversal to @xmath580 . by lemma [ homotimplieshomol ] ,
@xmath2028 so that @xmath2029=\big[\widetilde\upsilon\big({\left\langle}{\mathscr{k } } ' { \right\rangle},{\left\langle}{\mathscr{l}}^1{\right\rangle}\big)\big]$ ] . by lemma [ upsilon_tilde_def ] , the polycycles
@xmath1654 and @xmath2030 are homologous so that @xmath2031=\left[{\mathscr{d}}\left({\mathscr{k}},{\mathscr{l}}^1\right)\right]$ ] .
thus , in order to prove , we may assume without loss of generality that @xmath675 is transversal to @xmath1286 .
we need to prove that @xmath2032=\left[{\mathscr{d}}({\mathscr{k}},{\mathscr{l}})\right].\ ] ] set @xmath2033 , @xmath2034 and @xmath2035 it follows from the definition that @xmath2036 .
thus @xmath1813 is obtained by cutting @xmath56 into pieces , each piece " being a connected component of @xmath2037 where @xmath534 is a @xmath102-simplex of @xmath469 . the map @xmath2038 is the obvious gluing map .
it is surjective and its restriction to every connected component of @xmath1813 is injective .
consider the equivalence relations @xmath2039 on @xmath56 and @xmath2040 on @xmath1813 defined by the partitions ( see section [ partitions ] ) .
we claim that if some points @xmath2041 satisfy @xmath2042 , then @xmath2043 .
we now check this claim .
assume that for @xmath1175 , @xmath2044 by assumption , there exist faces @xmath2045 of @xmath56 of the same type such that @xmath2046 for @xmath1175 and @xmath2047 carries @xmath2048 to @xmath2049 .
for @xmath1175 , let @xmath2050 and @xmath2051 be faces of @xmath409 and @xmath337 respectively such that @xmath2052 is the smallest face of @xmath1824 containing @xmath497 .
then @xmath2053 , @xmath2054 , @xmath2055 has the same type as @xmath2056 , @xmath2057 has the same type as @xmath2058 , and @xmath2059 , @xmath2060 . to proceed , let @xmath2061 be the connected component of @xmath2062 in @xmath560 , and let @xmath2063 be the corresponding singular simplex ( which is a simplicial isomorphism onto a @xmath102-simplex of the triangulation @xmath469 )
. then @xmath2064 . by the assumption ( * )
, @xmath2065 is a face of the @xmath102-simplex @xmath2066 .
since @xmath469 fits @xmath443 , the sets @xmath2067 are faces of the @xmath102-simplices @xmath2068 containing @xmath2069 respectively . the map @xmath2070 restricts to a simplicial isomorphism @xmath2071 preserving the order of the vertices and carrying @xmath195 to @xmath196 . set @xmath2072
then @xmath2073 is an @xmath418-dimensional face of the @xmath102-simplex @xmath2074 containing @xmath2062 for @xmath1175 .
the map @xmath2075 is an order - preserving simplicial isomorphism and @xmath2076 .
since @xmath2077 are copies of the standard @xmath102-simplex @xmath1063 , their faces @xmath2078 correspond to certain @xmath2079-element subsets @xmath2080 of the set @xmath2081 . since @xmath2082
, we have @xmath2083 by the definition of @xmath1286 , the latter equality implies that the faces @xmath2084 , @xmath2085 of @xmath560 have the same type and @xmath2086 .
consider now the map @xmath2087 we have @xmath2088 let @xmath2089 be the connected component of @xmath2090 in @xmath2091 .
the equality @xmath2092 implies that @xmath2093 .
thus , @xmath2094 and @xmath2095 are faces of @xmath1813 of the same type , and the identification map @xmath2096 carries @xmath2097 to @xmath2098 .
this proves that @xmath2043 as claimed .
consider the canonical projections @xmath2099 and @xmath2100 .
the previous claim implies that there exists a unique map @xmath1299 such that the following diagram commutes : @xmath2101_-g & \ar[l]_-\pi d \\
d'_{\theta ' } & d ' \ar[u ] _ { \overline \zeta\vert_{d ' } } \ar[l]_-{\pi'}. } \ ] ] the map @xmath1299 is continuous because @xmath2102 is continuous and @xmath2103 are quotient maps .
set @xmath2104 where @xmath2105 is the @xmath2106-skeleton of @xmath469 .
clearly , @xmath2107 .
consider the commutative diagram @xmath2108 & \ar[l]_-{\pi _ * } h_*(d,\partial d ) \ar[d ] & \!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\ ! \!\!\!\!\ ! \!\!\!\!\!\ni [ d , w ] \\
[ d_\theta , w]\in \!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\ ! & h_*\left(d_\theta \right ) \ar[r ] \ar[d]_-{g _ * } \ar[ru ] & h_*\left(d_\theta , s_\theta \right ) \ar[d]_-{g _ * } & \ar[l]_-{\pi _ * } h_*(d , s ) & \!\!\!\!\ ! \!\!\!\!\ ! \!\!\!\!\!\!\!\ !
\ni ( \overline \zeta\vert _ { d ' } ) _ * ( [ d',w ' ] ) \\ \big[d'_{\theta'},w'\big]\in \!\!\!\!\!\!\!\!\!\!\!\!\!\ ! & h_*\left(d'_{\theta'}\right ) \ar[r ] & h_*\left(d'_{\theta'},(\partial d')_{\theta'}\right ) & \ar[l]_-{\pi ' _ * } h_*(d',\partial d ' ) \ar[u]_-{(\overline \zeta\vert _ { d ' } ) _ * } & \!\!\!\!\!\!\!\!\!\!\!\!\!\ ! \!\!\!\!\ ! \!\!\!\!\!\ni [ d',w ' ] } \ ] ] where the unlabelled arrows are the inclusion maps .
the definition of @xmath1286 implies that the weight @xmath2109 of @xmath1286 is the composition of @xmath2110 with the weight @xmath2111 of @xmath580 . using the definition of the operation @xmath1591
, we deduce that the weight @xmath2112 is the composition of @xmath2113 with the weight @xmath2114 . this fact and the definition of @xmath2115 , [ d , w]$ ]
imply that @xmath2116\right)$ ] is the image of @xmath2117 $ ] in @xmath2118 . by lemma [ partit ] , the image of @xmath2119
$ ] in @xmath2120 is equal to @xmath2121)$ ] . using this fact , the uniqueness in lemma [ partit ] , and the diagram chasing we obtain that @xmath2122 ) = [ d'_{\theta'},w ' ] \ \in h_\ast(d'_{\theta'}).\ ] ] next , we verify that @xmath2123 . for a point @xmath2124
, we have @xmath2125 for any @xmath2126 .
then @xmath2127 where we use the equality @xmath2128 .
since @xmath2099 is onto , we conclude that @xmath2129 . this equality and imply : @xmath2130 & = & \left((\kappa { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda ) _ { \theta}\right ) _ * \left([d_{\theta},w]\right)\\ & = & \left((\kappa ' { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda ) _ { \theta'}\right ) _ * g_*([d_\theta , w])\\ & = & \left((\kappa ' { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda ) _ { \theta'}\right ) _ * \left([d'_{\theta'},w']\right ) \ = \
\left[{\mathscr{d}}({\mathscr{k}}',{\mathscr{l}})\right].\end{aligned}\ ] ]
we introduce the path homology category of a manifold and define the intersection bibracket in this category . throughout this section ,
@xmath0 is a smooth oriented connected @xmath405-dimensional manifold with non - void boundary and @xmath1509 .
let @xmath2131 be the graded category whose set of objects is @xmath92 and whose graded modules of morphisms are defined by @xmath2132 for any @xmath2133 .
composition in @xmath350 is the pairing defined through concatenation of paths . for @xmath2134 ,
the identity morphism of @xmath2135 in @xmath350 is the element of @xmath2136 represented by the constant path in @xmath2137 .
we call @xmath350 the _ path homology category _ of @xmath0 . by section [ extenscat ] , this category determines a graded algebra @xmath2138 the subcategory @xmath2139 of @xmath350 formed by all objects and morphisms of degree @xmath183 can be formulated in terms of paths in @xmath0 : for any @xmath2133 , the module of morphisms @xmath2140 is freely generated by the set of homotopy classes of paths from @xmath2137 to @xmath2141 in @xmath0 .
thus the category @xmath2139 is the linearization of the fundamental groupoid @xmath2142 of @xmath0 based at @xmath92 , and the algebra @xmath2143 is the corresponding groupoid algebra .
clearly , @xmath2143 embeds in @xmath7 as a subalgebra .
assume that @xmath2144 and @xmath2145 by the k " unneth theorem , the condition holds if @xmath95 is a principal ideal domain and @xmath2146 is a flat @xmath95-module ( this occurs , for instance , when @xmath95 is a field ) ; it also holds for any @xmath95 if @xmath2147 is a free abelian group .
then the cross product induces an isomorphism @xmath2148 for any choice of base points @xmath2149 . composing the inverse isomorphism with the map @xmath2150 defined in section [ upsilon1111 ]
, we obtain a linear map @xmath2151 the direct sum of these maps over all 4-tuples of points in @xmath3 is a linear map @xmath2152 called the _ intersection bibracket _ of @xmath0 .
we can now state our main result .
[ geometricbibracket ] under the assumptions above , @xmath385 is a double gerstenhaber category of degree @xmath2153 . that the bibracket @xmath6 has degree @xmath107 follows from the fact that the intersection polychain of a @xmath102-polycycle and a @xmath711-polycycle has dimension @xmath2154 for any @xmath233 .
the @xmath107-antisymmetry of @xmath2155 follows from the formula and the following well - known fact : for any topological spaces @xmath81 and @xmath683 such that the cross product induces an isomorphism @xmath2156 , the isomorphism @xmath2157 induced by the interchange of factors @xmath2158 and the cross product isomorphisms , carries @xmath2159 to @xmath2160 for any homogeneous @xmath2161 , @xmath2162 ; see , for example , ( * ? ? ?
* section 4(b ) ) . that the bibracket @xmath2155 satisfies the first @xmath107-graded leibniz rule follows from lemma [ leibniz_upsilon ] using the associativity and the naturality of the cross product in singular homology . by lemma [ bibrtranspose ] and the @xmath107-antisymmetry
, the bibracket @xmath6 also satisfies the second @xmath107-graded leibniz rule .
therefore @xmath6 is a @xmath107-graded bibracket in @xmath7 .
it is obvious from the definitions that @xmath6 annihilates the identity morphisms of all objects .
it remains only to prove that the induced tribracket is equal to zero ; we postpone the proof to section [ jacobi ] . since @xmath2163 ,
the restriction of the intersection bibracket in @xmath350 to @xmath2139 is equal to zero .
moreover , the morphisms in @xmath2139 represented by paths in @xmath92 annihilate the bibracket in @xmath350 both on the right and on the left .
theorem [ geometricbibracket ] and lemma [ bracketcategories ] imply that for every integer @xmath16 , the associated representation algebra @xmath2164 has a natural structure of a gerstenhaber algebra of degree @xmath2165 .
we now specialize the results of section [ interbi ] by fixing a base point @xmath2166 . by the _ pontryagin algebra of @xmath0 _ ,
we mean the unital graded algebra @xmath2167 multiplication in @xmath2168 is the _ pontryagin product _ given by @xmath2169 .
pontryagin algebras have been extensively studied since serre s thesis @xcite .
they can be explicitly computed using the adams hilton model @xcite or the techniques of rational homotopy theory ( at least , in the simply connected case ) .
we only mention the relation with the homotopy groups , and refer to @xcite for a detailed exposition . consider the boundary homomorphism @xmath2170 for the path space fibration of @xmath0 where @xmath2171 and @xmath2172 is the constant path at @xmath60 . since the total space of that fibration is contractible , @xmath2173 is an isomorphism for all @xmath496 .
composing @xmath2173 with the hurewicz homomorphism @xmath2174 , we obtain an additive map @xmath2175 called the _
connecting homomorphism_. for @xmath2176 , this homomorphism extends to a ring isomorphism @xmath2177 \simeq a_\star^0 $ ] . for @xmath2178 ,
this homomorphism induces an isomorphism from @xmath2179 onto @xmath2180 where @xmath2181 is the connected component of @xmath2182 formed by all null - homotopic loops .
the group @xmath2183 acts on @xmath2168 by graded algebra automorphisms : the action of any @xmath2184 is the automorphism @xmath2185 of @xmath2168 where @xmath1299 is viewed as an invertible element of @xmath2186 .
the inclusion @xmath2187 allows us to consider the induced action of @xmath2188 on @xmath2168 .
[ geometricbibracket+++++ ] if @xmath2189 and the condition is satisfied , then the restriction of the intersection bibracket @xmath6 in the category @xmath2190 to @xmath2168 is a @xmath2188-equivariant gerstenhaber bibracket of degree @xmath2165 . clearly , @xmath2191 where @xmath2192 is the full subcategory of @xmath350 determined by the object @xmath60 .
therefore our claim is a consequence of theorem [ geometricbibracket ] and the results stated at the end of section [ doublecat ] .
we only need to check the equivariance .
let @xmath2193 and let @xmath61 be a loop in @xmath92 based at @xmath60 representing @xmath2194 .
we deduce from the commutativity of the diagram that @xmath2195 using the naturally of the cross product , we conclude that @xmath2196 by theorem [ geometricbibracket+++++ ] and lemma [ importantgerspp ] , the intersection bibracket in @xmath2168 induces a natural structure of a gerstenhaber algebra of degree @xmath2197 in the commutative unital graded algebra @xmath2198 for all @xmath16 .
we call @xmath2198 the _ @xmath34-th representation algebra of @xmath0_. the action of @xmath2188 on @xmath2168 induces an action of @xmath2188 on @xmath2198 by graded algebra automorphisms , and the gerstenhaber bracket in @xmath2198 is @xmath2188-equivariant
. the isomorphism classes of the double gerstenhaber algebra @xmath2168 and the gerstenhaber algebras @xmath2199 depend only on the connected component of @xmath60 in @xmath92 .
we keep notation of section [ the pontryagin algebra ] and let @xmath2200 be the quotient of @xmath2201 by the submodule @xmath2202 $ ] spanned by the vectors @xmath2203 where @xmath2204 run over all homogeneous elements of @xmath2205 . under the assumptions of theorem [ geometricbibracket+++++ ] ,
the intersection bibracket @xmath6 in @xmath2168 composed with the multiplication of @xmath2168 induces a @xmath2206-graded lie bracket @xmath2207 in @xmath2200 , see section [ stdb ] .
this bracket can be computed using the map @xmath2208 induced by the concatenation of loops .
namely , if @xmath2209 is the natural projection , then for any homogeneous @xmath2210 , @xmath2211\big).\end{aligned}\ ] ] the resulting expression may be used as the definition of the bracket @xmath2207 avoiding the use of @xmath6 .
this gives a @xmath2206-graded lie bracket in @xmath2212 over an arbitrary commutative ring @xmath95 .
the jacobi identity for @xmath2207 may be deduced from lemma [ jacobi_upsilon ] below .
presumably , the bracket @xmath2213 is related to the operation discussed in ( * ? ? ?
* remark 3.2.3 ) using chas
sullivan s techniques .
suppose that the manifold @xmath0 is simply connected and the ground ring @xmath95 is a field of characteristic zero .
the classical milnor moore theorem ( see ( * ? ? ? * theorem 21.5 ) ) asserts that , the pontryagin algebra @xmath2201 is fully determined by @xmath2214 and the whitehead bracket @xmath2215_{\operatorname{wh}}$ ] in @xmath2216 .
more precisely , consider the graded module @xmath2217 ( obtained from @xmath2216 by tensorizing with @xmath95 and shifting the degree by @xmath300 ) , and equip @xmath2218 with the bracket defined by @xmath2219= kl \otimes ( -1)^{p + 1 } \partial_{p+q+1 } \left ( \left[\partial_{p+1}^{-1 } ( \alpha ) , \partial_{q+1}^{-1 } ( \beta)\right ] _ { \operatorname{wh } } \right ) \
\in { \mathbb{k}}\otimes \pi_{p+q } ( \omega_\star)\ ] ] for any @xmath2220 , @xmath2221 , @xmath2222 .
then @xmath2223 is a @xmath183-graded lie algebra and the hurewicz homomorphism @xmath2224 extends to an isomorphism of the universal enveloping algebra @xmath2225 onto @xmath2205 . moreover , under this isomorphism , the standard comultiplication in @xmath2226 carrying any @xmath2227 to @xmath2228 corresponds to the comultiplication in @xmath2168 induced by the diagonal map @xmath2229 .
note that , by the poincar
witt theorem for graded lie algebras ( * ? ? ? * theorem 21.1 ) , the natural linear map @xmath2230 is injective so that @xmath2218 can be treated as a submodule of @xmath2231 .
recall from section [ amt1 ] that the @xmath183-graded lie algebra @xmath2232 gives rise to representation algebras @xmath2233 .
the milnor
moore isomorphism @xmath2234 induces an isomorphism @xmath2235 for all @xmath16 . in this way
, the algebras @xmath2233 acquire a structure of gerstenhaber algebras of degree @xmath2165 .
we briefly discuss the case @xmath1970 which , in fact , gave the original impetus to this work .
a connected oriented surface @xmath0 with @xmath2236 is an eilenberg
maclane space @xmath2237 where @xmath2238 is the fundamental group of @xmath575 . for any points @xmath2239 ,
the space @xmath2240 is homotopy equivalent to the underlying discrete set of @xmath11 . therefore ,
in the notation of section [ phc ] , we have @xmath2241 and @xmath2242 is the groupoid algebra of @xmath2142 . for any points @xmath2243
such that @xmath2244 , section [ upsilon1111 ] yields a linear map @xmath2245 ( the latter equality holds for all @xmath95 ) .
this construction extends to arbitrary 4-tuples of points in @xmath92 by slightly pushing these points in the positive direction along @xmath92 and proceeding as in section [ checkupsilon ] . after an appropriate normalization ,
this yields a @xmath183-antisymmetric 0-graded bibracket of degree 0 in the groupoid algebra @xmath7 .
this bibracket is quasi - poisson in an appropriate sense , cf@xmath134 @xcite . for @xmath47 , the restriction of this bibracket to the group algebra
@xmath2246 $ ] is the double bracket @xmath2247 studied in ( * ? ? ?
* section 7 ) .
it is closely related to the homotopy intersection form in @xmath2248 $ ] introduced in @xcite ; see also @xcite for a similar operation .
the associated lie bracket @xmath2207 in @xmath2200 was first introduced by goldman @xcite .
lemma [ bracketcategories ] implies that for every integer @xmath16 , the above bibracket in @xmath7 induces a ( quasi - poisson ) bracket in the associated representation algebra @xmath2164 .
note that @xmath2249 is the coordinate algebra of the affine scheme ( over @xmath95 ) that associates to any unital commutative algebra @xmath24 the set of groupoid homomorphisms @xmath2250 .
indeed , through linear extension of groupoid homomorphisms , the latter set may be identified with the set of linear functors from @xmath2251 to the category @xmath150 considered in section [ extenscat ] .
we conclude by applying .
we conclude the proof of theorem [ geometricbibracket ] by proving that the tribracket associated with the intersection bibracket is equal to zero . we keep notation of section [ notat ] and
begin by developing a parametrized version of the theory of polychains in path spaces .
let @xmath1057 be an arbitrary topological space . given a polychain @xmath712 in @xmath2252 , we let @xmath2253 and @xmath2254 be the compositions of @xmath2255 with the projections to @xmath2256 and @xmath2257 , respectively .
we call the polychain @xmath675 _ smooth _ if the map @xmath2258 is smooth in the sense of section [ path homology ] . applying the definitions of section [ facehomology ] but considering only smooth polychains in @xmath2252
, we obtain _ smooth face homology _ @xmath2259 .
the proof of theorem [ loopsp+++++ ] easily adapts to this setting and yields that the natural linear map @xmath2260 is an isomorphism .
this computes the face homology of @xmath2261 in terms of smooth polychains in @xmath2252 .
we say that smooth polychains @xmath1219 in @xmath1593 and @xmath1461 in @xmath2262 are if the maps @xmath2263 and @xmath2264 are transversal in the sense of section [ transs ] .
a pair @xmath2265 with @xmath1673 is _ transversely represented _ by a pair @xmath2266 if @xmath580 is a smooth reduced @xmath102-polycycle in @xmath1652 and @xmath2267 is a smooth reduced @xmath711-polycycle in @xmath2268 transversal to @xmath580 . adapting the proof of lemma [ keytheor ] , we obtain that any pair @xmath1757 as above can be transversely represented by a pair of polycycles , and , furthermore , any two such pairs of polycycles can be related by a finite sequence of transformations @xmath1477 of the following types : * @xmath1478 and @xmath1479 or @xmath1480 where @xmath968 is a smooth @xmath1658-polychain in @xmath1652 transversal to @xmath675 ; * @xmath1481 and @xmath1482 or @xmath1483 where @xmath1484 is a smooth @xmath1662-polychain in @xmath2269 transversal to @xmath580 .
we next adapt the construction of the intersection polychain .
consider smooth transversal polychains @xmath634 in @xmath1593 and @xmath712 in @xmath2252 . since the polychain @xmath2270 in @xmath1595 is smooth and transversal to @xmath580 , section [ def_d ] yields an intersection polychain @xmath2271 in @xmath1584 .
we lift @xmath2272 to a polychain in @xmath2273 as follows .
let @xmath2274 be the cartesian projection .
the tuple @xmath2275 with @xmath2276 is a polychain in @xmath2273 .
we need only to check that the map @xmath2277 is compatible with the partition @xmath1580 .
let @xmath449 be two faces of @xmath56 of the same type and let @xmath2278 be the smallest faces of @xmath1824 containing @xmath449 , respectively .
since @xmath2279 is compatible with the partition @xmath853 of @xmath337 , we have for any @xmath2280 @xmath2281 the next lemma is a parametrized version of lemma [ upsilon_tilde_def ] and is proved similarly .
[ upsilon^x_tilde_def ] for any integers @xmath1646 , the intersection @xmath2282 induces a bilinear map @xmath2283 the direct sum over all integers @xmath1673 of the pairings produced by lemma [ upsilon^x_tilde_def ] is a linear map of degree @xmath1674 @xmath2284 as in section [ checkupsilon ] , a normalized version of this map @xmath2285 is defined by @xmath2286 for any homogeneous @xmath1964 and @xmath2287 .
we also define an operation @xmath2288 in singular homology by the commutative diagram @xmath2289 ^ - { \check\upsilon_{12,34{z } } } & & \widetilde h_{\ast } ( \omega_{32 } \times \omega_{14 } \times { z } ) \ar[d]^-{[-]}\\ h_\ast ( \omega_{12 } ) \otimes h_\ast ( \omega_{34 } \times { z } ) \ar[u]^-{{\left\langle}-{\right\rangle}\times { \left\langle}-{\right\rangle } } \ar@{-->}[rr]^- { \upsilon_{12,34{z } } } & & h_{\ast } ( \omega_{32 } \times \omega_{14 } \times { z } ) . } \ ] ] the proof of lemma [ diag - imp ] extends to this new setting and gives the commutative diagram @xmath2290 ^ - { \check\upsilon_{12,34{z } } } \ar[d]_-{[- ] \times [ - ] } & & \widetilde h_{\ast } ( \omega_{32 } \times \omega_{14 } \times { z } ) \ar[d]^-{[-]}\\ h_\ast ( \omega_{12 } ) \otimes h_\ast ( \omega_{34 } \times { z } ) \ar[rr]^- { \upsilon_{12,34{z } } } & & h_{\ast } ( \omega_{32 } \times \omega_{14 } \times { z } ) . } \ ] ] the following two lemmas will help us to compute @xmath2291 and @xmath2292 .
[ tilde_upsilon_cross ] for any @xmath2293 , any @xmath2294 and any homogeneous @xmath2295 , we have @xmath2296 it suffices to consider the case where @xmath32 and @xmath122 are homogeneous .
let @xmath634 and @xmath712 be smooth polycycles in @xmath1652 and @xmath1653 representing @xmath32 and @xmath122 respectively , and such that @xmath580 is transversal to @xmath675 .
let @xmath2297 be a polycycle in @xmath2298 representing @xmath2299 .
then @xmath2300 by the definition of @xmath2301 , @xmath2302 therefore , it is enough to show that @xmath2303 we set @xmath2304 so that @xmath2305 we also set @xmath2306 where @xmath2307 and @xmath2308 are the cartesian projections . the map @xmath2309 restricts to a diffeomorphism @xmath2310 of degree @xmath2311 . for any point @xmath2312 in @xmath2313
, we have @xmath2314 furthermore , the diffeomorphism @xmath259 carries the partition @xmath2315 to @xmath2316 and the weight @xmath2317 to the weight @xmath2318 .
hence , @xmath259 is a diffeomorphism of polychains .
[ upsilon_cross ] for any @xmath2319 , @xmath2320 and @xmath2321 , we have @xmath2322 it is enough to prove the lemma for homogeneous @xmath293 . by lemma [ twocrossproducts ] , we have @xmath2323 \times [ { \left\langle}c { \right\rangle}]= [ { \left\langle}b { \right\rangle}\times { \left\langle}c{\right\rangle}]$ ] .
we deduce that @xmath2324\otimes [ { \left\langle}b { \right\rangle}\times { \left\langle}c{\right\rangle}]\big ) \\ & = & \left[\check { \upsilon}_{12,34{z}}\big({\left\langle}a{\right\rangle}\otimes ( { \left\langle}b { \right\rangle}\times { \left\langle}c { \right\rangle } ) \big ) \right ] \\ & = & ( -1)^{\vert b \vert + \vert c \vert + n \vert a \vert } \big[\widetilde{\upsilon}_{12,34{z}}\big({\left\langle}a{\right\rangle}\otimes ( { \left\langle}b { \right\rangle}\times { \left\langle}c { \right\rangle } ) \big ) \big]\\ & = & ( -1)^{\vert b \vert + n \vert a \vert } \big[\widetilde{\upsilon}_{12,34{}}({\left\langle}a{\right\rangle}\otimes { \left\langle}b { \right\rangle } ) \times { \left\langle}c { \right\rangle}\big ] \\ & = & ( -1)^{\vert b \vert + n \vert a \vert } \big[\widetilde{\upsilon}_{12,34 { } } ( { \left\langle}a{\right\rangle}\otimes { \left\langle}b { \right\rangle } ) \big ] \times \left [ { \left\langle}c { \right\rangle}\right ] \\ & = & { \upsilon}_{12,34 { } } \big(a\otimes b\big ) \times c\end{aligned}\ ] ] where the second , fourth and fifth equalities follow from , lemma [ tilde_upsilon_cross ] and lemma [ twocrossproducts ] respectively . for further use , we generalize the results above to the following situation .
assume that we have two topological spaces @xmath683 and @xmath1057 .
a straightforward generalization of lemma [ upsilon^x_tilde_def ] yields a bilinear map @xmath2325 an appropriate normalized version @xmath2326 of this map is defined by @xmath2327 for any homogeneous @xmath2328 and @xmath2329 .
the corresponding map in singular homology is defined by the commutative diagram @xmath2330 ^ - { \check\upsilon_{{y}12,34{z } } } & & \widetilde h_{\ast } ( y \times \omega_{32 } \times \omega_{14 } \times { z } ) \ar[d]^-{[-]}\\ h_\ast ( y \times \omega_{12 } ) \otimes h_\ast ( \omega_{34 } \times { z } ) \ar[u]^-{{\left\langle}-{\right\rangle}\times { \left\langle}-{\right\rangle } } \ar@{-->}[rr]^- { \upsilon_{{y}12,34{z } } } & & h_{\ast } ( y \times \omega_{32 } \times \omega_{14 } \times { z } ) . } \ ] ] then , again , we have the commutative diagram @xmath2331 ^ - { \check\upsilon_{{y}12,34{z } } } \ar[d]_-{[- ] \times [ - ] } & & \widetilde h_{\ast } ( y \times \omega_{32 } \times \omega_{14 } \times { z } ) \ar[d]^-{[-]}\\ h_\ast ( y \times \omega_{12 } ) \otimes h_\ast ( \omega_{34 } \times { z } ) \ar[rr]^- { \upsilon_{{y}12,34{z } } } & & h_{\ast } ( y \times \omega_{32 } \times \omega_{14 } \times { z } ) . } \ ] ] finally , lemma [ upsilon_cross ] generalizes to the identity @xmath2332 for any @xmath2319 , @xmath2320 and @xmath2333 , @xmath2334 .
we compute the intersection operations of section [ parar ] via so - called `` half - smooth '' polychains .
let @xmath1057 be an arbitrary topological space .
a @xmath711-polychain @xmath2335 is _ half - smooth _ if the restrictions of the map @xmath2336 ( adjoint to @xmath2337 ) to the manifolds with faces @xmath2338 $ ] and @xmath2339 $ ] are smooth .
furthermore , @xmath675 is _ half - transversal _ to a smooth @xmath102-polychain @xmath594 in @xmath1593 if for any face @xmath1363 of @xmath409 , any face @xmath439 of @xmath337 , and any of the three sets @xmath2340 , [ 1/2,1 ] , \{1/2\}$ ] the map @xmath2341 is weakly transversal to @xmath2342 in the sense of section [ transs ] .
then the set @xmath2343 inherits a structure of a manifold with faces from @xmath2344 , and we have @xmath2345 in the sequel we denote @xmath2346 ) , \quad d^+=d([1/2 , 1 ] ) , \quad d^{\nicefrac{1}{2}}=d(\{1/2\}).\ ] ] it is clear that @xmath2347 and @xmath2348 since @xmath675 may be non - smooth , we can not consider the intersection polychain @xmath2349 .
( a priori , the set @xmath2350 does not have a structure of a manifold with faces . ) instead , we turn the disjoint union @xmath2351 into a polychain which will serve as a substitute for @xmath2349 .
the inclusion allows us to use the same construction as in section [ def_d ] in order to upgrade @xmath2352 , @xmath2353 , and @xmath2354 to polychains in @xmath2355 denoted , respectively , @xmath2356 , @xmath2357 , and @xmath2358 .
as can be checked from our conventions , the oriented manifold @xmath2354 has the orientation inherited from @xmath2359 or , equivalently , the orientation inherited from @xmath2360 .
the inclusions @xmath2361 are compatible with the polychain structures ( except for the orientations ) : they map faces of @xmath2362 diffeomorphically onto faces of @xmath2363 , map faces of the same type onto faces of the same type , commute with the identification diffeomorphisms of the faces , commute with the maps to @xmath2355 , and the induced maps in @xmath2364 commute with the weights .
moreover , a face of @xmath2363 having the same type as the image of a face @xmath439 of @xmath2362 must be the image of a face of @xmath2362 of the same type as @xmath439 .
these facts allow us to form a @xmath1526-polychain @xmath2365 in @xmath2355 by taking the disjoint union @xmath2366 and declaring that the images of any face of @xmath2354 in @xmath2367 and @xmath2368 have the same type and the identification diffeomorphism between them is the identity map .
we shall sometimes write @xmath2369 for this polychain @xmath2370 .
[ half - half ] let @xmath580 be a smooth @xmath102-polycycle in @xmath1593 and let @xmath675 be a half - smooth @xmath711-polycycle in @xmath2252 half - transversal to @xmath580 . then @xmath2371 is a polycycle in @xmath2372 and @xmath2373 = [ { \mathscr{d}}^{h } ( { \mathscr{k}},{\mathscr{l } } ) ] \in h_{p+q+2-n}(\omega_{32 } \times \omega_{14 } \times { z})$ ] .
lemma [ reduction_d ] directly extends to smooth polychains @xmath580 , @xmath1286 in @xmath1593 and half - smooth polychains @xmath1594 in @xmath2252 half - transversal to @xmath580 , @xmath1286 ; one should only replace @xmath1591 by @xmath2374 .
this implies the first claim of lemma [ half - half ] .
there is an arbitrarily small deformation @xmath2375 of @xmath2376 into a smooth polycycle @xmath2027 .
moreover , we can assume that the restrictions of the maps @xmath2377 \to m$ ] to @xmath2378 $ ] and @xmath2339 $ ] are smooth maps smoothly depending on @xmath689 .
as in the proof of lemma [ homotimplieshomol ] , we derive from the deformation @xmath2379 a @xmath1662-polychain @xmath955 in @xmath2268 such that @xmath2380 .
the assumptions on the deformation imply that @xmath955 is half - smooth .
taking the deformation small enough , we can ensure that @xmath955 is half - transversal to @xmath580 . by the assumption @xmath2381 and the generalized version of lemma [ reduction_d ] , @xmath2382
therefore @xmath2383 projecting to singular homology , we obtain the equality @xmath2384= \left [ { \mathscr{d}}^{h } ( { \mathscr{k}},{\mathscr{l}}^1)\right ] \in { h}_{p+q+2-n}(\omega_{32 } \times \omega_{14 } \times { z}).\ ] ] since the polycycle @xmath2027 is smooth , the manifold with faces underlying @xmath2385 is obtained by cutting out the manifold with faces underlying @xmath2386 along a smooth compact oriented proper submanifold of codimension 1 .
this easily implies the equality @xmath2387= \left[{\mathscr{d}}^{z}({\mathscr{k}},{\mathscr{l}}^1)\right ] $ ] .
thus , @xmath2388 & = & \big[\widetilde{\upsilon}_{12,34z}({\left\langle}{\mathscr{k}}{\right\rangle } , { \left\langle}{\mathscr{l}}^1{\right\rangle } ) \big ] \\ & = & \left[{\mathscr{d}}^{z}({\mathscr{k}},{\mathscr{l}}^1)\right ] \ = \ \left
[ { \mathscr{d}}^{h } ( { \mathscr{k}},{\mathscr{l}})\right].\end{aligned}\ ] ] assume that @xmath2189 .
as in section [ checkupsilon ] , the operations @xmath2301 , @xmath2389 , @xmath2390 generalize to arbitrary tuples of points @xmath2391 in @xmath92 .
we pick two extra points @xmath2392 . for @xmath2393 , the maps @xmath2301 , @xmath2389 , @xmath2390
will be denoted respectively by @xmath2394 , @xmath2395 , @xmath2396 . given a permutation @xmath2397 of @xmath2398 , we can accordingly renumber the points @xmath2399 and consider the corresponding maps @xmath2400 , @xmath2401 , @xmath2402 .
we now establish a jacobi - type identity for @xmath2403 .
[ jacobi_upsilon ] [ jacobi_upsilon_tilde ] consider the permutation maps @xmath2404 for any @xmath2405 , @xmath2406 and @xmath2407 with @xmath2408 , we have the following equality in @xmath2409 : @xmath2410 set @xmath2411 .
the definition of @xmath2412 and imply that @xmath2413 \otimes \left[\check { \upsilon}_{34,56}({\left\langle}b { \right\rangle}\otimes { \left\langle}c { \right\rangle})\right]\right ) \\ & = & \left [ \check{\upsilon}_{12,5436 } \left({\left\langle}a { \right\rangle}\otimes \check { \upsilon}_{34,56}({\left\langle}b { \right\rangle}\otimes { \left\langle}c { \right\rangle } ) \right ) \right ] \\ &
= & ( -1)^{(r+nq)+ ( q+r+n + np ) } \big [ \widetilde{\upsilon}_{12,5436 } \big({\left\langle}a { \right\rangle}\otimes \widetilde{\upsilon}_{34,56}({\left\langle}b { \right\rangle}\otimes { \left\langle}c { \right\rangle } ) \big ) \big]\\ & = & \varepsilon ( -1)^{q+p(n+r ) } \big [ \widetilde{\upsilon}_{12,5436 } \big({\left\langle}a { \right\rangle}\otimes \widetilde{\upsilon}_{34,56}({\left\langle}b { \right\rangle}\otimes { \left\langle}c { \right\rangle } ) \big ) \big].\end{aligned}\ ] ] using the naturality of the transformation @xmath774 $ ] , we also obtain that @xmath2414 \\ & = & ( -1)^{(p+n ) ( q+r)+(p+nr)+(r+p+n + nq ) } \big[({\mathsf{p}}_{312 } ) _ * \widetilde{\upsilon}_{34,1652 } \big({\left\langle}b { \right\rangle}\otimes \widetilde{\upsilon}_{56,12}({\left\langle}c { \right\rangle}\otimes { \left\langle}a{\right\rangle})\big)\big]\\ & = & \varepsilon ( -1)^{r+q(n+p ) } \big[({\mathsf{p}}_{312 } ) _ * \widetilde{\upsilon}_{34,1652 } \big({\left\langle}b { \right\rangle}\otimes \widetilde{\upsilon}_{56,12}({\left\langle}c { \right\rangle}\otimes { \left\langle}a{\right\rangle})\big)\big]\end{aligned}\ ] ] and @xmath2415 \\ & = & ( -1)^{(p+q)(r+n ) + ( q+np)+ ( p+q+n + nr ) } \big[({\mathsf{p}}_{231 } ) _ * \widetilde{\upsilon}_{56,3214 } \big({\left\langle}c { \right\rangle}\otimes \widetilde{\upsilon}_{12,34}({\left\langle}a { \right\rangle}\otimes { \left\langle}b { \right\rangle})\big ) \big ] \\ & = & \varepsilon ( -1)^{p+r(n+q ) } \big[({\mathsf{p}}_{231 } ) _ * \widetilde{\upsilon}_{56,3214 } \big({\left\langle}c { \right\rangle}\otimes \widetilde { \upsilon}_{12,34}({\left\langle}a { \right\rangle}\otimes { \left\langle}b { \right\rangle})\big ) \big].\end{aligned}\ ] ] thus , it is enough to prove the following identity in @xmath2416 , where @xmath1768 , @xmath1769 and @xmath2417 are now any face homology classes : @xmath2418&&\\ \notag + ( -1)^{r+q(n+p)}\ , \big[({\mathsf{p}}_{312 } ) _
* \widetilde{\upsilon}_{34,1652 } \big(b \otimes \widetilde{\upsilon}_{56,12}(c \otimes a)\big ) \big]&&\\ \notag + ( -1)^{p+r(n+q)}\ , \big [ ( { \mathsf{p}}_{231 } ) _ * \widetilde{\upsilon}_{56,3214 } \big(c \otimes \widetilde{\upsilon}_{12,34}(a \otimes b)\big ) \big ] & = & 0 .\end{aligned}\ ] ] slightly moving the points @xmath2399 in @xmath92 , we can assume that they are pairwise distinct .
let @xmath634 be a smooth @xmath102-polycycle in @xmath1593 representing @xmath32 , let @xmath712 be a smooth @xmath711-polycycle in @xmath1595 representing @xmath122 , and let @xmath2297 be a smooth @xmath418-polycycle in @xmath2419 representing @xmath2299 .
we will assume that @xmath2420 are pairwise transversal in the sense of section [ transs ] .
this assumption and other transversality conditions imposed below in the course of the proof are always achieved by a small deformation of @xmath2421 .
let @xmath2422 be the intersection polycycle as defined in section [ def_d ] . recall that its underlying manifold with faces , @xmath2423 , consists of all tuples @xmath2424 such that @xmath2425 .
let @xmath2426 stand for the underlying continuous map @xmath2427 of @xmath2428 .
the map @xmath2429 carries a point @xmath2430 to the path @xmath2431 which runs from @xmath2432 to @xmath2433 along @xmath2434 in the first half - time and then runs from @xmath2435 to @xmath2436 along @xmath2437 in the second half - time .
( here and below , the time parameter of paths always increases along subintervals of @xmath644 with constant speed . )
the map @xmath2438 carries @xmath2430 to the path @xmath2439 which runs from @xmath2440 to @xmath2441 along @xmath2437 in the first half - time and then runs from @xmath2442 to @xmath2443 along @xmath2444 in the second half - time .
thus the paths @xmath2445 and @xmath2446 are obtained from the paths @xmath2447 and @xmath2437 by switching direction at the intersection point @xmath2425 , see figure [ dbc ] .
[ r ] at 2 215 [ r ] at 2 121 [ r ] at 2 24 [ l ] at 195 21 [ l ] at 195 121 [ l ] at 195 216 [ b ] at 57 180 [ b ] at 46 101 [ b ] at 149 118 [ b ] at 143 36 in @xmath2448.,title="fig : " ] we set @xmath2449 , @xmath2450 and view @xmath2428 as a polycycle in @xmath2451 .
it is half - smooth in the sense of section [ parar+++ ] . slightly deforming the map @xmath2452 adjoint to @xmath73
, we can assume @xmath2428 to be half - transversal to @xmath580 in the sense of section [ parar+++ ] .
in the sequel , we consider the associated @xmath2453-polychains @xmath2454 and @xmath2455 in @xmath2456 . on the one hand ,
the manifold with faces @xmath2457 underlying the polychain @xmath2458 consists of all tuples @xmath2459\ ] ] such that @xmath2460 and @xmath2461 .
the map @xmath2462 underlying @xmath2458 is schematically shown in figure [ dabc ] where one switches direction at the dotted intersections .
the first coordinate @xmath2463 sends any point to the path @xmath2464 which goes from @xmath2432 to @xmath2465 along @xmath2434 in half - time and , next , goes from @xmath2466 to @xmath1949 along @xmath2467 in half - time .
the map @xmath2468 carries a point to the path @xmath2431 which goes from @xmath1948 to @xmath2469 along @xmath2470 in half - time , next , goes from @xmath2465 to @xmath2471 along @xmath2434 in time @xmath2472 $ ] and , finally , goes from @xmath2435 to @xmath2436 along @xmath2437 in time @xmath2473 $ ] .
the map @xmath2474 sends a point to the path @xmath2475 which goes from @xmath2440 to @xmath2476 along @xmath2477 in half - time and , next , goes from @xmath2478 to @xmath2443 along @xmath2444 in half - time . on the other hand ,
the manifold with faces @xmath2479 underlying the polychain @xmath2480 consists of all tuples @xmath2481 such that @xmath2460 and @xmath2482 .
the map @xmath2483 is computed similarly to and is schematically shown in figure [ dabc ] .
we only note that the map @xmath2484 carries a point to the path @xmath2431 which goes from @xmath2432 to @xmath2471 along @xmath2434 in time @xmath2485 $ ] , next , goes from @xmath2486 to @xmath2487 along @xmath2477 in time @xmath2488 $ ] and , finally , goes from @xmath2466 to @xmath1949 along @xmath2467 in the remaining half - time .
consider also the polychain @xmath2489 in @xmath2456 .
its underlying @xmath2490-manifold with faces @xmath2491 consists of the tuples @xmath2492 such that @xmath2493 .
the underlying map @xmath2494 is the restriction of the maps and , see figure [ dabc ] .
[ r ] at 5 213 [ r ] at 292 218 [ r ] at 586 221 [ r ] at 5 120 [ r ] at 292 123 [ r ] at 586 122 [ r ] at 5 22 [ r ] at 292 25 [ r ] at 586 24 [ l ] at 193 215 [ l ] at 193 119 [ l ] at 192 19 [ l ] at 480 215 [ l ] at 480 119 [ l ] at 480 19 [ l ] at 773 215 [ l ] at 773 119 [ l ] at 773 21 [ b ] at 29 199 [ b ] at 128 193 [ b ] at 25 42 [ bl ] at 85 125 [ b ] at 140 110 [ b ] at 142 38 [ b ] at 20 112 [ b ] at 344 180 [ b ] at 435 36 [ b ] at 323 105 [ b ] at 438 115 [ b ] at 327 50 [ b ] at 424 202 [ b ] at 633 185 [ b ] at 730 36 [ b ] at 629 100 [ b ] at 699 106 [ b ] at 749 122 [ b ] at 627 35 [ b ] at 731 190 , @xmath2495 and @xmath2480.,title="fig : " ] cyclically permuting @xmath293 , we similarly obtain polychains @xmath2496 , @xmath2497 and @xmath2498 , @xmath2499 .
lemma [ half - half ] allows us to rewrite as the identity @xmath2500\\
\notag + ( -1)^{r+q(n+p)}\ , ( { \mathsf{p}}_{312 } ) _ * \big [ { \mathscr{d}}_{bca}^{-}\mathop{\cup}_{\nicefrac{1}{2 } } { \mathscr{d}}_{bca}^{+ } \big]&&\\ \notag + ( -1)^{p+r(n+q)}\ , ( { \mathsf{p}}_{231 } ) _ * \big [ { \mathscr{d}}_{cab}^{- } \mathop{\cup}_{\nicefrac{1}{2 } } { \mathscr{d}}_{cab}^{+ } \big ] & = & 0\end{aligned}\ ] ] in @xmath2501 .
the idea of the proof is to show that the six polychains on the left - hand side of cancel each other pairwise .
we first explain how to relate the polychains @xmath2502 and @xmath2503 .
observe that the manifold with faces @xmath2504 underlying @xmath2503 consists of all tuples @xmath2505 such that @xmath2506 and @xmath2507 .
we define a smooth map
@xmath2508 \to l \times i^\circ \times n \times i^\circ \times k \times i^\circ \times [ 1/2,1)\ ] ] by the formula @xmath2509 where @xmath2510 \to i^\circ$ ] and @xmath2511 \to [ 1/2,1)$ ] are given by @xmath2512 observe that the functions @xmath2513 satisfy the equations @xmath2514 and @xmath2515 .
it easily follows that the transformation @xmath2516
\to i^\circ \times [ 1/2,1)$ ] is a diffeomorphism , so that @xmath439 is a diffeomorphism carrying @xmath2457 onto @xmath2504 .
the resulting diffeomorphism @xmath2517 is compatible with the partitions and the weights of the polychains @xmath2518 , @xmath2503 .
moreover , @xmath2519 up to homotopy of the second coordinate map compatible with the partitions . the map @xmath2520 carries @xmath2521 diffeomorphically onto @xmath2522 via the permutation @xmath2523 and holds on @xmath2524 as an equality of maps ( no homotopy needed ) .
one easily constructs a homotopy of the map @xmath2525 into @xmath2526 constant on @xmath2527 .
since the left - hand side of is preserved under such a homotopy of @xmath2528 , we can assume that is an equality of maps .
we prove now that @xmath2529 the diffeomorphism @xmath2530 carries the open subset @xmath2531 of @xmath2457 onto the open subset @xmath2532 of @xmath2533 , and @xmath2534 is equal to the degree of the restricted diffeomorphism @xmath2535 . clearly , @xmath2536 and @xmath2537 consider the maps @xmath2538 and @xmath2539 since @xmath1484 is transversal to both @xmath580 and @xmath675 , the map @xmath2540 is transversal to @xmath2541 in the following sense : for any faces @xmath2542 of @xmath2543 respectively , the restriction of @xmath2540 to the interior of @xmath2544 is transversal to the interior of @xmath2541 ( in the usual sense of differential topology ) .
similarly , the map @xmath2545 is transversal to @xmath2541 .
observe that @xmath2546 and that the inverse images of @xmath2541 under the maps @xmath2547 are , respectively , the sets @xmath2548 .
we identify @xmath2549 where @xmath2550 is the cartesian projection defined by @xmath2551 .
as above , @xmath2552 carries the orientation induced by that of @xmath2553 using our orientation convention , and we give to the product orientation .
pulling back the latter orientation along @xmath2540 , we obtain an orientation on the normal bundle of @xmath2554 in @xmath2555 ; this oriented normal vector bundle is denoted by @xmath2556 .
the normal bundle of @xmath2557 in @xmath2558 is oriented similarly and denoted by @xmath2559 .
let @xmath2560 be the tangent bundle of @xmath2554 with the orientation induced by that of @xmath2556 .
similarly , let @xmath2561 be the tangent bundle of @xmath2557 with the orientation induced by that of @xmath2559 .
clearly , the diffeomorphism @xmath2516
\to i^\circ \times [ 1/2,1)$ ] defined by is orientation - reversing .
hence @xmath2562 , and since @xmath439 carries @xmath2563 onto @xmath2557 and induces an orientation - preserving map @xmath2564 , we have @xmath2565 next , consider the following isomorphisms of oriented vector bundles over @xmath2554 , where @xmath469 stands for the tangent bundle , @xmath1632 stands for the normal bundle , and @xmath2566 denotes the appropriate cartesian projection : @xmath2567 it follows that @xmath2568 .
similarly , @xmath2569 here the sign @xmath1643 accompanying @xmath2559 is the degree of the permutation map @xmath2570 .
it follows that @xmath2571 .
formula and the computations of @xmath2572 imply .
cyclically permuting @xmath293 , we obtain diffeomorphisms @xmath2573 and @xmath2574 such that @xmath2575 and @xmath2576 to conclude the proof , we set @xmath2577 clearly , @xmath2578 on @xmath2527
. therefore any triangulation of @xmath2527 extends uniquely to a triangulation , @xmath2579 , of @xmath2580 invariant under @xmath2581 .
( all triangulations in this argument are supposed to be locally ordered and to fit the given partitions , cf@xmath134 sections [ partitions ] and [ fundcla ] . )
subdividing , if necessary , @xmath2579 we can assume that it extends to a triangulation , @xmath2560 , of @xmath2582 .
transferring @xmath2560 along the diffeomorphism @xmath2583 we obtain a triangulation , @xmath2561 , of @xmath2353 also extending @xmath2579 .
we use the triangulations @xmath2560 and @xmath2561 to represent the left - hand side of by a @xmath2453-dimensional singular chain . according to , and , every singular simplex contributed by a top - dimensional simplex of @xmath2560 cancels with the corresponding singular simplex in @xmath2561 .
therefore the singular chain in question is equal to zero and so is the left - hand side of .
let @xmath2584 be the tribracket induced by the intersection bibracket @xmath6 in @xmath2585 .
pick any points @xmath2586 and any homology classes @xmath2405 , @xmath2406 and @xmath2407 .
we need to show that the tensor @xmath2587 vanishes , where @xmath2588 are the graded permutations defined in section [ conventions ] . for any @xmath2589 ,
let @xmath2590 @xmath2591 be the linear maps induced by the cross product . by definition of the intersection bibracket and lemma [ upsilon_cross ] , @xmath2592 cyclically permuting @xmath293 , we also obtain @xmath2593 and @xmath2594 combining the last three identities , formula and lemma [ jacobi_upsilon ] , we obtain that @xmath2595 .
we conclude that @xmath2596 .
we compute @xmath1975 for spherical homology classes of complementary dimensions and for @xmath183-dimensional homology classes .
we use these results to determine the intersection bibracket in two examples .
let @xmath0 be a smooth connected oriented manifold of dimension @xmath2597 with non - empty boundary .
we compute the operation @xmath1975 on the loop homology classes arising from spheres of complementary dimensions .
let us fix a base point @xmath2598 in the @xmath189-sphere @xmath2599 for every @xmath113 . for @xmath2600 , we let @xmath2601 $ ] be the @xmath189-th homotopy group of @xmath0 at @xmath209 . for @xmath2602 , we set @xmath2603 .
consider base points @xmath2604 in @xmath92 and integers @xmath2605 such that @xmath2606 .
let @xmath2607 be the following composition : @xmath2608_-{\overline \partial_p \times \overline \partial_q } \ar@/^0.5cm/@{-->}[drr]^-{\upsilon^\pi_{\star,\star ' } } & & \\
h_{p-1}(\omega_\star ) \otimes h_{q-1}(\omega_{\star ' } ) \ar[r]^-\upsilon & h_0(\omega_{\star'\!\star } \times \omega_{\star\star ' } ) \simeq \hspace{-1 cm } & \ { \mathbb{k}}[\pi_1(m,\star',\star ) ] \otimes { \mathbb{k}}[\pi_1(m,\star,\star ' ) ] . } } \ ] ] here @xmath2609 , @xmath2610 , @xmath2611 is the connecting homomorphism of section [ the pontryagin algebra ] , and similar notation applies with @xmath60 and @xmath63 exchanged .
the following lemma computes @xmath2607 when @xmath2612 .
[ intersecting_spheres ] assume @xmath2613 .
let @xmath2614 and @xmath2615 be continuous maps such that @xmath2616 , @xmath2617 and @xmath2618 , @xmath2619 are transversal smooth maps
. then @xmath2620,[\beta ] ) = ( -1)^{n(p+1)+1 } \sum_{(x , y ) } \varepsilon(x , y ) \ , [ \beta _ { y } \alpha_{x } ^{-1 } ] \otimes [ \alpha _ { x } \beta_{y } ^{-1 } ] .\ ] ] here : the sum runs over all @xmath2621 such that @xmath2622 ; @xmath2623 is the sign of the product orientation in @xmath2624 with respect to the orientation of @xmath0 ; @xmath2625 is the composition of @xmath207 with a path from @xmath2626 to @xmath209 in @xmath2627 and @xmath2628 is the composition of @xmath2629 with a path from @xmath2630 to @xmath1040 in @xmath2631 . for @xmath113 ,
let @xmath2632 be a continuous map such that @xmath2633 , @xmath2634 is smooth and the quotient map @xmath2635 is a degree 1 homeomorphism .
then @xmath2636 is adjoint to a continuous map @xmath2637 which carries @xmath2638 to the constant path @xmath2639 .
let @xmath2640 be the quotient map .
then @xmath2641\right ) = ( \bar \omega_\alpha)_*\big(\big[i^{p-1}/\partial i^{p-1}\big]\big ) = \left[{\mathscr{k}}\right]\ ] ] where @xmath2642 is the polycycle in @xmath2182 with weight @xmath300 and with partition @xmath2643 defined as the product of @xmath2644 copies of the partition of @xmath644 identifying @xmath645 to @xmath646 .
similarly , @xmath2645\right ) = \left[{\mathscr{l}}\right]$ ] for @xmath2646 .
the polycycles @xmath580 and @xmath675 are admissible in the sense of section [ computationoftildeupsilon ] where @xmath2647 , @xmath2648 , @xmath2649 and @xmath2650
. we can therefore consider the intersection polychain @xmath1654 and by lemma [ newcomputation ] , it represents @xmath2651 .
then , using lemma [ diag - imp ] , we get @xmath2652 , [ \beta ] ) \ = \ \upsilon([{\mathscr{k } } ] , [ { \mathscr{l } } ] ) & = & ( -1)^{(q-1)+n(p-1 ) } \big[\widetilde \upsilon\left(\ , { \left\langle}{\mathscr{k}}{\right\rangle}\otimes { \left\langle}{\mathscr{l}}{\right\rangle}\ , \right ) \big]\\ & = & ( -1)^{q+n(p+1)+1 } \big [ { \mathscr{d } } ( { \mathscr{k } } , { \mathscr{l } } ) \big ] .\end{aligned}\ ] ] the intersection polycycle @xmath2653 is 0-dimensional , and its points bijectively correspond to the pairs @xmath2621 such that @xmath2622 . such a pair @xmath2654 contributes @xmath2655 to @xmath2656 where @xmath2657 are paths as in the statement of the lemma , @xmath2658 is the composition of @xmath207 with a path from @xmath209 to @xmath2626 in @xmath2627 , and @xmath2659 is the composition of @xmath2629 with a path from @xmath1040 to @xmath2630 in @xmath2631 . here
@xmath2660 is the sign of the linear isomorphism @xmath2661 ^ -{(\alpha \times \beta ) _ * } & t_{(z , z ) } \left(m \times m \right)\ar[r ] & \frac{t_{(z , z ) } ( m \times m)}{t_{(z , z ) } { \operatorname{diag}}_m } = \nu_{m \times m}({\operatorname{diag}}_m)_{(z , z ) } , } \ ] ] where @xmath2662 , @xmath2663 has the product orientation and @xmath2664 has the orientation induced from that of @xmath2553 .
the linear map @xmath2665 defined by @xmath2666 induces an orientation - preserving isomorphism @xmath2667 . composing with the linear isomorphism above
, we obtain the map @xmath2668 whose degree is @xmath2669 .
therefore @xmath2670 .
thus , @xmath2671 , [ \beta ] ) & = & ( -1)^{n(p+1)+1 } \sum_{(x , y ) } \varepsilon(x , y)\big [ \left ( \beta _ { y } \bar \alpha_{x } , \alpha _ { x } \bar\beta_{y } \right ) \big]\\ & = & ( -1)^{n(p+1)+1 } \sum_{(x , y ) } \varepsilon(x , y ) \big [ \beta _ { y } \bar \alpha_{x } \big ] \otimes \big [ \alpha _ { x } \bar \beta_{y } \big].\end{aligned}\ ] ] since @xmath2672 , the path @xmath2673 is well defined up to homotopy rel @xmath2674 and @xmath2675 is homotopic to @xmath2676 .
similar claims hold for @xmath2629 since @xmath2677 . this yields .
if we consider a single point @xmath60 in the boundary of @xmath575 , then we can similarly compute the linear map @xmath2678 \otimes { \mathbb{k}}[\pi_1(m,\star ) ] .\ ] ] fix a path @xmath2679 from @xmath52 to a different point @xmath2680 , and consider maps @xmath2614 and @xmath2615 satisfying the conditions of lemma [ intersecting_spheres ] . transporting @xmath2681 along @xmath72
, we obtain a map @xmath2682 . applying lemmas [ leibniz_upsilon ] and [ intersecting_spheres ]
, we obtain that @xmath2683 , [ \varsigma^{-1 } \beta ] ) & = & \upsilon\!\left ( \overline \partial_p [ \alpha ] , \varsigma ( \overline \partial_q [ \beta ] ) \varsigma^{-1 } \right ) \\ \notag & = & \varsigma\ , \upsilon\!\left ( \overline \partial_p [ \alpha ] , \overline \partial_q [ \beta]\right ) \varsigma^{-1 } \\
\notag & = & \varsigma\ , \upsilon^\pi_{\star,\star'}([\alpha ] , [ \beta ] ) \varsigma^{-1 } \\ \label{upsilon_pi } & = & ( -1)^{n(p+1)+1 } \sum_{(x , y ) } \varepsilon(x , y ) \ , [ \varsigma \beta _ { y } \alpha_{x } ^{-1 } ] \otimes [ \alpha _ { x } \beta_{y } ^{-1}\varsigma^{-1 } ] .\end{aligned}\ ] ] this computation implies that the map @xmath2684 is determined by the pairing @xmath2685 , \ ] ] where @xmath2686 \to { \mathbb{k}}$ ] is the addition of coefficients .
note that @xmath2687 , [ \varsigma^{-1 } \beta ] ) = ( -1)^{n(p+1)+1 } \sum_{(x , y ) } \varepsilon(x , y ) \ , [ \alpha _ { x } \beta_{y } ^{-1}\varsigma^{-1 } ] .\ ] ] the pairing on the right - hand side is the well known geometric intersection `` of spherical cycles or the reidemeister pairing '' , see @xcite or ( * ? ? ?
* section 5 ) .
let @xmath0 be a smooth connected oriented manifold of dimension @xmath1921 with non - empty boundary .
we fix three points @xmath2688 and consider the map @xmath2689 defined by the following composition : @xmath2690_- { \overline \partial_1 \times \overline \partial_{n-1 } } \ar@/^0.5cm/@{-->}[drr]^-{\upsilon^\pi_{12,3 } } & & \\
h_{0}(\omega_{12 } ) \otimes h_{n-2}(\omega_{33 } ) \ar[r]^-\upsilon & h_0(\omega_{32 } \times \omega_{13 } ) \simeq \hspace{-1 cm } & \ { \mathbb{k}}\big[\pi_1(m,\star_3,\star_2)\big ] \otimes { \mathbb{k}}\big[\pi_1(m,\star_1,\star_3)\big ] . } } \ ] ] as in the previous sections , @xmath2691 for any @xmath2692 .
lemma [ intersecting_spheres ] easily adapts to this setting and yields the following computation of @xmath2693 .
[ intersecting_arc_and_spheres ] let @xmath2694 and let @xmath2695 be a continuous map such that @xmath2696 .
assume that @xmath2697 , @xmath2698 and that @xmath2699 , @xmath2700 are transversal smooth maps
. then @xmath2701,[\beta ] ) = - \sum_{(x , y ) } \varepsilon(x , y ) \ , [ \beta _ { y } \alpha_{x 1 } ] \otimes [ \alpha _ { 0x } \beta_{y } ^{-1 } ] .\ ] ] here : the sum runs over all @xmath2702 \times s^{n-1}$ ] such that @xmath2622 ;
@xmath2623 is the sign of the product orientation in @xmath2703 ) \oplus \beta_*(t_y s^{n-1})=t_{\alpha ( x ) } m$ ] with respect to the orientation of @xmath0 ; @xmath2704 ( respectively @xmath2705 ) is the path running along @xmath207 from @xmath1948 to @xmath2706 ( respectively from @xmath2706 to @xmath1949 ) in the positive direction and @xmath2628 is the composition of @xmath2629 with a path from @xmath2707 to @xmath1040 in @xmath2708 .
lemma [ intersecting_arc_and_spheres ] can be adapted to the cases where @xmath2709 and/or @xmath2710 .
besides , we can similarly define an operation @xmath2711 \otimes { \mathbb{k}}\big[\pi_1(m,\star_1,\star_3)\big]\ ] ] and compute it as in lemma [ intersecting_arc_and_spheres ] .
fix @xmath2712 integers @xmath2713 such that @xmath2714 .
consider the closed smooth @xmath405-manifold @xmath2715 with the product orientation on each summand , and set @xmath2716 where @xmath2717 is a closed @xmath405-ball smoothly embedded in @xmath2718 .
fix a point @xmath47 and consider the pontryagin algebra @xmath2719 where @xmath2609 .
we now compute the intersection bibracket in @xmath2168 .
for an appropriate choice of @xmath2717 , of the base points @xmath2720 , and of the balls along which the connected sums are performed in , the sets @xmath2721 are embedded spheres in @xmath0 , for all @xmath2722 .
since @xmath0 is simply connected , these spheres define certain elements @xmath2723 and @xmath2724 , respectively .
consider the corresponding elements of the pontryagin algebra @xmath2725 since @xmath0 deformation retracts to a wedge of @xmath2712 spheres isotopic to @xmath2726 , it follows from ( * ? ? ?
* iii.1.b ) ( or , alternatively , from ( * ? ? ?
* corollary 2.2 ) ) that @xmath2727 @xmath2728 @xmath2729 freely generate the unital graded algebra @xmath2168 . in particular ,
if @xmath763 , then @xmath2168 is a free abelian group .
therefore the condition is satisfied for any ground ring @xmath95 .
hence the intersection bibracket @xmath6 in @xmath2168 is defined for any @xmath95 , and is fully determined by its values on the generators .
these values can be computed from the formula : for any @xmath2730 , @xmath2731 @xmath2732 here we use the assumption that the spheres @xmath2733 have codimension @xmath2734 in @xmath0 and so can be made disjoint from the interiors of arcs connecting them to @xmath60 . as a consequence , we observe that the bibracket @xmath6 is a graded version of the bibracket associated by van den bergh @xcite with the ( double of the ) quiver @xmath2735 having a single vertex and @xmath1299 edges .
the graded module @xmath2736 $ ] is freely generated by words in the letters @xmath2737 , subject to the _ cyclic relations _
@xmath2738 for any words @xmath2739 where @xmath2740 is the sum of the degrees of the letters appearing in @xmath2741 .
the @xmath2206-graded lie bracket @xmath2742 in @xmath2200 induced by @xmath6 is a graded version of the necklace lie bracket associated to @xmath2735 , see @xcite . for any integer @xmath16 ,
the gerstenhaber bracket @xmath2743 in @xmath2744 induced by @xmath6 can be computed from , . in particular
, @xmath2745 is the unital commutative graded algebra with free generators @xmath2737 in degrees @xmath2746 , and for any @xmath2747 , @xmath2748 the bracket @xmath2743 is a graded version of the standard poisson bracket in the symmetric algebra of a free module of rank @xmath2712 equipped with a symplectic form .
let @xmath2749 .
we compute the intersection bibracket in the pontryagin algebra of the exterior of a ball in @xmath2750 .
we endow @xmath2751 with the product orientation and set @xmath2752 where @xmath2753 and @xmath2754 are the base points .
set @xmath2755 where @xmath2717 is a closed @xmath405-ball smoothly embedded in @xmath2756 .
pick a point @xmath2757 and connect it to the point @xmath2758 by a path @xmath2759 such that @xmath2760 . up to homotopy relative to the endpoints , there are two such paths ; we take the path @xmath2761 such that a positive tangent vector of @xmath2761 followed by a positively oriented basis of @xmath2762 yields a positively oriented basis of @xmath2763 , see figure [ s1sq ] . transporting @xmath81 and @xmath683 along @xmath2761
, we obtain certain homotopy classes @xmath2764 and @xmath2765 .
consider the corresponding elements @xmath2766 of the algebra @xmath2719 .
note that @xmath209 is invertible in @xmath2767 $ ] .
we claim that the unital graded algebra @xmath2168 is generated by @xmath2768 and @xmath1040 subject to the only relation @xmath2769 .
indeed , @xmath2770 where @xmath2181 is the connected component of @xmath2182 consisting of null - homotopic loops .
the space @xmath2181 can be identified with the loop space of the universal cover of @xmath0 .
this cover has the homotopy type of a wedge of countably many copies of @xmath2708 since @xmath0 deformation retracts to @xmath2771 .
therefore , the unital graded algebra @xmath2772 is freely generated by the elements @xmath2773 , and the claim above easily follows .
[ ] at 149 113 [ ] at 147 22 [ ] at 152 70 [ t ] at 263 114 [ ] at 370 217 [ ] at 370 157 [ ] at 370 100 [ t ] at 665 143 [ tr ] at 92 29 .,title="fig : " ] in particular , if @xmath763 , then @xmath2168 is a free abelian group . therefore the intersection bibracket @xmath6 in @xmath2168
is defined for any ground ring @xmath95 . to determine @xmath6 , it suffices to compute its values on the generators @xmath2774 . for degree reasons ,
@xmath2775 let @xmath61 be an arc in @xmath92 connecting @xmath60 to another point @xmath63 . by lemma [ intersecting_arc_and_spheres ]
, we obtain @xmath2776 this implies the equalities @xmath2777 observe next that @xmath2778 and @xmath2779 are images under the connecting homomorphism of certain elements of @xmath2780 and @xmath2781 that can be represented by disjoint embedded spheres .
it follows that @xmath2782 which implies that @xmath2783 . using the leibniz rules and
, we deduce that @xmath2784 using , one can also compute the graded lie bracket @xmath2207 in @xmath2200 and the gersthenhaber bracket @xmath2743 in @xmath2785 for any integer @xmath16 .
we derive from the intersection bibracket a scalar intersection form and compute it in terms of usual homology intersections in manifolds .
consider an arbitrary graded category @xmath2786 and the associated graded algebra @xmath2787 , see section [ extenscat ] .
an _ augmentation _ of @xmath2786 is a degree - preserving linear map @xmath2788 mapping all the identity morphisms to @xmath300 and such that @xmath2789 for any morphisms @xmath2790 in @xmath2791 such that the target object of @xmath259 coincides with the source object of @xmath1299 . the condition that @xmath1328 is degree - preserving means that @xmath2792 for all @xmath2793 .
given an augmentation @xmath2788 and a @xmath107-graded bibracket @xmath6 in @xmath2786 with @xmath106 , we define the _ induced scalar form _ @xmath2794 by @xmath2795 for any @xmath178 .
observe that @xmath2796 for any @xmath28 , @xmath2797 , @xmath2798 with @xmath2799 ; similarly , @xmath2800 for any @xmath2801 , @xmath2802 , @xmath2803 with @xmath2799 .
furthermore , if the bibracket @xmath6 is @xmath107-antisymmetric , then @xmath2804 for any homogeneous @xmath178 . consider a smooth oriented connected manifold @xmath0 of dimension @xmath1509 with non - empty boundary and assume that condition is satisfied .
the path homology category @xmath2805 has a canonical augmentation @xmath2806 obtained as the direct sum over all @xmath2807 of the compositions @xmath2808 where the left arrow is the obvious projection and the right arrow carries the homology classes of all points to @xmath300 . by the previous paragraph , this augmentation together with the intersection bibracket induce a bilinear form @xmath2809 .
we compute @xmath2810 in terms of standard homological intersections in @xmath0 . for simplicity , we assume in the rest of this section that @xmath2749 , though the case @xmath1970 may be considered similarly . for any points @xmath2811
, we define a linear map @xmath2812 it suffices to define the restriction of @xmath2813 to @xmath2814 for any @xmath2815 . if @xmath2816 , then this restriction is equal to zero .
suppose now that @xmath2817 .
when @xmath2818 , the form @xmath2813 is the standard homological intersection , see , for example , @xcite .
when @xmath2819 , we separate two cases . if @xmath2820 , then @xmath2821 is canonically isomorphic to @xmath2822 and the pairing @xmath2813 is induced by the homological intersection @xmath2823 .
the case @xmath2824 is treated similarly using that @xmath2825 is canonically isomorphic to @xmath2826 .
note that the assumption @xmath2827 guarantees that @xmath2828 or @xmath2829 .
if both these inequalities hold true , then the two definitions above give the same pairing .
the next lemma yields a version of the homological suspension homomorphism due to serre @xcite .
[ serre ] let @xmath2830 with @xmath1237 .
there is a unique homomorphism @xmath2831 such that for every polycycle @xmath634 in @xmath1212 , we have @xmath2832)= \big[(k \times i , \varphi \times \tau , u \times 1 , \tilde\kappa ) \big]\ ] ] where @xmath486 is the trivial partition on @xmath643 $ ] . the uniqueness of @xmath1422 is a direct consequence of theorem [ mainonpolychains ] . to prove the existence ,
define a continuous map @xmath2833 by @xmath2834 and set @xmath2835 . the formula @xmath2836 , \\ \alpha(s + 2t-1 ) & \hbox{if } t \in\left[1/2-s/{2},1-s/{2 } \right ] , \\
\star ' & \hbox{if } t\in\left[1-s/{2 } , 1\right ] \end{array}\right.\ ] ] defines a continuous map @xmath2837 .
let @xmath2838 be the linear map sending any @xmath2839 to @xmath2840 ) $ ] ( the definition of @xmath534 is inspired by @xcite and ( * ? ? ?
* remark 3.2.3 ) ) . finally , we set @xmath2841 . to check ,
observe that the fundamental class @xmath2842\in h_1(i,\partial i)$ ] is represented by the @xmath300-dimensional polycycle @xmath2843 relative to @xmath2674 .
lemma [ twocrossproducts ] implies that for any polycycle @xmath634 in @xmath87 , @xmath2844 ) & = & { \operatorname{ev } } _ * d _ * ( [ { \mathscr{k } } ] \times [ \mathscr{i } ] ) \\ & = & ( { \operatorname{ev}}d ) _ * \left [ { \mathscr{k}}\times \mathscr{i } \right ] = \big[(k \times i , \varphi \times \tau , u \times 1 , \tilde\kappa ) \big].\end{aligned}\ ] ] we can now state the main result of this section . [ bulletvscdot ] for any @xmath2845 , the following diagram commutes : @xmath2846 ^ -{\bullet } \ar[d]_-{- ( -1)^{n\vert-\vert } \sigma\ , \otimes\ , \sigma } & { \mathbb{k}}\ , .
\\ h_*\big(m , \{\star_1,\star_2\}\big ) \otimes h_*\big(m , \{\star_3,\star_4\}\big ) \ar@/_1cm/[ru]^- { { { \displaystyle \mathop { \ \ } ^{\hbox{$\centerdot$ } } } } } & } \ ] ] the proof of theorem [ bulletvscdot ] proceeds in three steps .
first we consider arbitrary disjoint subsets @xmath2847 , @xmath2848 of @xmath92 and the standard homology intersection form @xmath2849 we denote this form by @xmath2850 and compute it in terms of polycycles .
secondly , we relate @xmath2850 to the operation @xmath1975 . finally , we deduce theorem [ bulletvscdot ] . [ bullet ]
let @xmath2847 , @xmath2848 be disjoint subsets of @xmath92 .
let @xmath634 be a smooth @xmath102-polycycle in @xmath0 relative to @xmath2847 , let @xmath712 be a smooth @xmath711-polycycle in @xmath0 relative to @xmath2851 such that the map @xmath2852 is transversal to @xmath1317 in the sense of section [ transs ] .
let @xmath2853 and let @xmath2854 be the cartesian projection .
then @xmath56 is a manifold with faces and , for some orientation , partition @xmath1580 , and weight @xmath1694 on @xmath56 , the polychain @xmath2855 is a polycycle such that @xmath2856 \odot [ { \mathscr{l } } ] = ( -1)^{q(p+n ) } [ { \mathscr{d } } ] \in h_{p+q - n}(m).\ ] ] the transversality assumption ensures that @xmath56 inherits from @xmath709 a structure of a manifold with corners , see @xcite .
the same argument as at the beginning of section [ def_d ] shows that @xmath56 is a manifold with faces .
we orient @xmath56 so that the induced orientation of its normal bundle in @xmath2857 is the pull - back of the orientation of the normal bundle of @xmath2858 in @xmath2859 via @xmath2860 .
the partition @xmath1580 of @xmath56 is defined as follows : the faces of @xmath56 are the connected components of the intersections @xmath2861 where @xmath439 and @xmath12 range over faces of @xmath409 and @xmath337 respectively ; two such faces @xmath2862 and @xmath2863 are of the same type if @xmath2864 are of the same type , @xmath2865 are of the same type , and @xmath2866
. then @xmath2867 .
the weight @xmath1694 of @xmath56 carries a connected component @xmath1057 of @xmath56 to @xmath2868 where @xmath351 are connected components of @xmath1522 respectively , such that @xmath2869 .
then @xmath2855 is a polycycle satisfying .
we leave the general case of this claim to the reader and prove it only under the following assumptions : @xmath409 and @xmath337 are transversal compact oriented smooth submanifolds of @xmath0 such that @xmath2870 and @xmath2871 ; the partitions @xmath443 of @xmath409 and @xmath853 of @xmath337 are trivial ; the weights @xmath2872 and @xmath2873 send all connected components to @xmath2874 ; the maps @xmath2875 and @xmath2876 are the inclusions . under these assumptions
, we have @xmath825 \odot [ { \mathscr{l } } ] = [ k ] \odot [ l]= [ k \cap l]\ ] ] where @xmath2877\in h_*(m,\partial_-m)$ ] , @xmath2878\in h_*(m,\partial_+m)$ ] and @xmath2879\in h_*(m)$ ] are the fundamental classes , and @xmath2880 is oriented so that @xmath2881 ( this agrees with the orientation rule in @xcite ) . since @xmath2882 corresponds to @xmath2883 under the standard identification @xmath1537
, we need only to compare the orientation of @xmath56 with that of @xmath2884 .
note the following orientation - preserving isomorphisms of oriented vector bundles : @xmath2885 where @xmath2886 and @xmath2887 are the cartesian projections . restricting to @xmath2888
, we obtain @xmath2889 where @xmath102 is the identification diffeomorphism @xmath2890 . on the other hand , @xmath2891 since @xmath2892 as oriented vector bundles
, we deduce that @xmath2893 and follows .
[ upsilon_epsilon ] let , under the assumptions of lemma [ bullet ] , @xmath2894 and @xmath2895 .
let @xmath2896 be the composition of the augmentation @xmath2897 with the linear map @xmath2898 sending @xmath2874 to @xmath2899\in h_0(m)$ ] . then the following diagram commutes : @xmath2900 ^ -{\upsilon_{12,34 } } \ar[d]_-{- ( -1)^{n\vert-\vert } \sigma\ , \otimes\ , \sigma } & h_*(\omega_{32 } \times \omega_{14 } ) \ar[d]^- { \varepsilon_1 } \\
h_*(m,\partial_- m ) \otimes h_*(m,\partial_+ m ) \ar[r]^- { \odot } & h_*(m ) . } \ ] ] let @xmath2901 be the cartesian projection . clearly , the map @xmath2902 is homotopic to the constant map @xmath2903 so that , in homology , @xmath2904 .
pick now any @xmath2905 and @xmath2906 with @xmath1673 .
let @xmath634 be a smooth reduced @xmath102-polycycle in @xmath1652 and let @xmath712 be a smooth reduced @xmath711-polycycle in @xmath1653 transversely representing the pair of face homology classes @xmath2907 .
set @xmath1606 .
then @xmath2908 ) \\ & = & ( -1)^{q+np } { \operatorname{ev}}_*\big[(d,\theta , w,\kappa\triangleleft \lambda)\big ] \\ & = & ( -1)^{q+np } \big[(d,\theta , w , \tilde \kappa \circ { \operatorname{pr}}\vert_d)\big]\end{aligned}\ ] ] where @xmath2909 is the cartesian projection .
we deduce from lemmas [ serre ] and [ bullet ] that @xmath2910 \odot \big[(l \times i , \psi \times \tau , v \times 1 , \tilde\lambda ) \big ] \\ & = & ( -1)^{1+np+(q+1)(p+1+n ) } \big[(d,\theta , w , \tilde \kappa \circ { \operatorname{pr}}\vert_d)\big ] \\ & = & ( -1)^{1+q + ( q+1)(p+1+n ) } \varepsilon_1 \upsilon_{12,34}(a \otimes b ) \ = \
( -1)^ { ( q+1)(p+n ) } \varepsilon_1 \upsilon_{12,34}(a \otimes b).\end{aligned}\ ] ] since @xmath2911 is even if @xmath2912 and @xmath2913 otherwise , we obtain the claim of the lemma .
we can now complete the proof of theorem [ bulletvscdot ] . set @xmath2914 and @xmath2915 .
suppose first that @xmath2916 . the desired claim is obtained by combining the diagram in lemma [ upsilon_epsilon ] with the obvious diagram @xmath2917_- { \simeq } ^- { \varpi_{32,14 } } \ar[d]_- { \varepsilon_1 } & h_*(\omega_{32 } ) \otimes h_*(\omega_{14 } ) \ar[d]^- { \varepsilon \otimes \varepsilon } \\
h_*(m ) \ar[r]^- { } & { \mathbb{k}}}\ ] ] where the bottom horizontal arrow is the standard augmentation . to handle the case @xmath2918 ,
consider a smooth isotopy @xmath2919 of @xmath2920 which is constant outside of a small neighborhood of the points @xmath2921 and such that the point @xmath2922 lies in @xmath2923 for @xmath1175 .
the diffeomorphism @xmath2924 induces horizontal isomorphisms in the commutative diagram @xmath2925 ^ - { \simeq } \ar[d]_- { - ( -1)^{n\vert-\vert } \sigma } & h_*\big(\omega(m , \star'_1 , \star'_2)\big ) \ar[d]_- { - ( -1)^{n\vert-\vert } \sigma } \\
h_*\big(m , \{\star_1,\star_2\}\big ) \ar[r]^- { \simeq } & h_*\big(m , \{\star'_1,\star'_2\}\big ) \ , . }
\ ] ] note that the upper horizontal arrow coincides with the isomorphism @xmath2926 defined in section [ upsilon1111 ] where @xmath1912 is the path @xmath2927
. tensoring this diagram by the obvious commutative diagram @xmath2928 ^ - { { \operatorname{id } } } \ar[d]_- { \sigma } & h_*\big(\omega(m , \star_3 , \star_4)\big ) \ar[d]_- { \sigma } \\
h_*\big(m , \{\star_3,\star_4\}\big ) \ar[r]^- { { \operatorname{id } } } & h_*\big(m , \{\star_3,\star_4\}\big ) } \ ] ] we obtain a commutative diagram @xmath2846 ^ - { \simeq } \ar[d]_-{- ( -1)^{n\vert-\vert } \sigma\ , \otimes\ , \sigma } & h_*\big(\omega(m , \star'_1 , \star'_2)\big ) \otimes h_*\big(\omega(m , \star_3 , \star_4)\big ) \ar[d]_-{- ( -1)^{n\vert-\vert } \sigma\ , \otimes\ , \sigma } \\ h_*\big(m ,
\{\star_1,\star_2\}\big ) \otimes h_*\big(m , \{\star_3,\star_4\}\big ) \ar[r]^- { \simeq } & h_*\big(m , \{\star'_1,\star'_2\}\big ) \otimes h_*\big(m , \{\star_3,\star_4\}\big ) \ , . }
\ ] ] by the first part of the proof , we have the diagram in theorem [ bulletvscdot ] for the points @xmath2929 . combining it with the diagram above we obtain the required diagram .
indeed , according to , the upper line represents the scalar form @xmath2930 . in the bottom line
we obviously get @xmath2931 .
the pontryagin algebra of a space carries a natural comultiplication and a natural antipode , and becomes thus a hopf algebra .
similar comultiplication and antipode in the category of paths leads to a notion of a hopf category .
we use this notion to formulate a fundamental geometric property of the intersection bibracket called `` reducibility '' . consider a graded category @xmath2786 and the associated graded algebra @xmath2787 , see section [ extenscat ] . for an object @xmath1216 of @xmath2786 , we let @xmath2932 be the element represented by the identity morphism of @xmath81 . we view @xmath2933 as an algebra with multiplication defined by @xmath2934 for any homogeneous @xmath2935 . in the notation of section [ conventions ] , we have @xmath2936 a _ comultiplication _ in the graded category @xmath2786 is a degree - preserving algebra homomorphism @xmath2937 such that @xmath2938 and @xmath2939 for all @xmath2940 . as a consequence , @xmath2941 must carry @xmath2942 to @xmath2943 for any objects @xmath2944 of @xmath2786 .
the image of any @xmath2945 under @xmath2941 expands ( non - uniquely ) as a sum @xmath2946 where @xmath496 runs over a finite set and @xmath2947 are homogeneous elements of @xmath2948 .
we use sweedler s notation , i.e. , drop the index @xmath496 and the summation sign and write simply @xmath2949 . the condition that @xmath504 is degree - preserving means that @xmath2950 for any homogeneous @xmath2951 .
the condition that @xmath534 is an algebra homomorphism means that @xmath2952 for any homogeneous @xmath39 .
a _ counit _ for a comultiplication @xmath2953 is an augmentation @xmath2788 of @xmath2786 in the sense of section [ scalar_form ] such that @xmath2954 thus , if @xmath2953 has a counit @xmath1328 , then @xmath534 is a split injection with left inverses @xmath2955 and @xmath2956 . consider a comultiplication @xmath534 in @xmath2786 with counit @xmath1328 .
we define linear endomorphisms @xmath2957 of @xmath115 by @xmath2958 for all @xmath2959 and @xmath2960 .
an _ antipode _ in @xmath2786 is a degree - preserving linear map @xmath2961 mapping @xmath352 to @xmath2962 for any @xmath2963 and satisfying @xmath2964 for all @xmath158 .
it follows immediately that @xmath2965 for any @xmath2940 .
a tuple consisting of a graded category @xmath2966 , a comultiplication @xmath534 in @xmath2966 , a counit @xmath1328 in @xmath2966 , and an antipode @xmath1005 in @xmath2966 is called a _
hopf category_. when the category @xmath2786 has a single object , we recover the usual notion of a _ graded hopf algebra_. a hopf category is _ cocommutative _ if @xmath2967 and is _ involutive _ if @xmath1005 is an involution .
basic properties of hopf algebras ( see , for instance , ( * ? ? ?
* theorem iii.3.4 ) ) generalize to hopf categories .
we state here the properties used in the sequel .
[ properties_antipode ] the antipode @xmath1005 of a hopf category @xmath2786 is an antiendomorphism of the underlying algebra of @xmath2968 in the sense that , for any homogeneous @xmath39 , @xmath2969 also , @xmath1005 is an antiendomorphism of the underlying counital coalgebra of @xmath7 in the sense that , for any @xmath28 , @xmath2970 finally , the cocommutativity of @xmath2786 implies its involutivity , and the latter is equivalent to any of the following two properties : * for all @xmath28 , @xmath2971 ; * for all @xmath28 , @xmath2972 . in the proof
we will use the following notation .
recall that the algebra @xmath7 is linearly generated by morphisms in @xmath2786 . given two expressions linearly depending on one or several elements @xmath2973 of @xmath7 ,
we relate these expressions by the symbol @xmath2974 if they are equal for all @xmath2973 and this equality follows from the axioms of a hopf category whenever @xmath2975 are morphisms in @xmath2786 .
let @xmath563 be the module of degree - preserving linear maps @xmath2976 .
note that the comultiplication in @xmath7 induces a degree - preserving coassociative comultiplication in @xmath2977 carrying @xmath2159 with @xmath178 to @xmath2978 this comultiplication induces the _ convolution product _ @xmath1549 in @xmath563 by @xmath2979 for any @xmath2980 and any @xmath39 .
we define elements @xmath2981 of @xmath2982 by @xmath2983 and @xmath2984 for any homogeneous @xmath39 .
to prove the first claim of the lemma we must show that @xmath2985 . to this end
we define elements @xmath2986 of @xmath563 by @xmath2987 for any @xmath39 .
observe that @xmath2988 and @xmath2989 furthermore , @xmath2990 and @xmath2991 since @xmath1549 is an associative operation , we deduce that @xmath2992 we now verify that @xmath1005 is an antiendomorphism of the unital coalgebra @xmath2993 .
for this , we consider the module @xmath56 of degree - preserving linear maps @xmath2994 , and we equip it with the _ convolution product _ defined by @xmath2995 for any @xmath2996 and any @xmath28 .
let @xmath2997 be defined by @xmath2998 we must prove that @xmath2985 . observe that , for any @xmath28 , @xmath2999 and @xmath3000 furthermore , @xmath3001 and @xmath3002 using the associativity of @xmath1549 , we deduce that @xmath3003 also , @xmath1005 preverves the counit : for any @xmath28 , we have @xmath3004 we now prove the part of the lemma concerning the involutivity . if @xmath3005 , then the condition ( i ) is satisfied : @xmath3006 assume now that the condition ( i ) is met and consider the convolution product @xmath1549 in the module , @xmath1363 , of degree - preserving linear maps @xmath3007 .
for any @xmath158 , @xmath3008 thus , @xmath3009 .
it follows from the axioms of a hopf category that @xmath3010 and @xmath3011 for each @xmath3012 carrying the set @xmath3013 into itself for all @xmath2963 . applying this to @xmath3014 and to @xmath3015 and using the associativity of @xmath1549 , we obtain @xmath3016 this shows the equivalence between the involutivity and ( i ) ; the equivalence with ( ii ) is proved similarly .
finally , if @xmath7 is cocommutative , then the identity @xmath3017 implies ( i ) , so that @xmath7 is involutive .
we reformulate bibrackets in hopf categories in a way that helps to reveal their algebraic properties .
consider a hopf category @xmath3018 with associated graded algebra @xmath2787 and consider a @xmath107-graded bibracket @xmath3019 with @xmath106 .
we define a linear map @xmath3020 by @xmath3021 for any @xmath178 .
note that , if @xmath3022 and @xmath3023 with @xmath3024 , then @xmath3025 [ f ] for any @xmath3026 and any homogeneous @xmath3027 , we have @xmath3028 since both sides of the first identity are linear in @xmath122 and @xmath2299 , it suffices to consider the case where @xmath3029 and @xmath3030 for some objects @xmath3031 of @xmath2786 .
if @xmath3032 , then both sides of the identity are equal to zero .
if @xmath3033 , then @xmath3034 and @xmath3035 where the signs @xmath3036 are computed by @xmath3037 the second identity is proved similarly with the key case being the one where @xmath125 are morphisms in @xmath2786 and the target object of @xmath32 coincides with the source object of @xmath122
. then @xmath3038 and @xmath3039 where the signs @xmath3040 are computed by @xmath3041 the map @xmath3042 induces a bilinear pairing @xmath3043 by the formula @xmath3044 . here is an explicit expression for this pairing : @xmath3045 it follows from lemma [ f ] that , for any @xmath158 and any homogeneous @xmath3046 , @xmath3047 note that @xmath3048 is the scalar form induced by @xmath6 , see section [ scalar_form ] .
the bibracket @xmath6 may be recovered from the map @xmath3049 at least in the case where the antipode @xmath1005 in @xmath2786 is invertible .
indeed , for any @xmath178 , @xmath3050 if follows that , if the antipode @xmath1005 is invertible , then @xmath3051 we introduce a class of bibrackets that may be recovered from the associated pairing @xmath74 .
we call a bibracket @xmath2155 _ reducible _ if @xmath3052
. then @xmath3053 as a consequence , a reducible bibracket in a hopf category with invertible antipode is fully determined by @xmath74 . as we explain below , reducible bibrackets naturally arise in the geometric setting .
the following lemma yields further properties of @xmath2155 , @xmath3054 , and @xmath1320 in the case of cocommutative @xmath3055 . [ fcoc ] suppose that the hopf category @xmath3055 is cocommutative .
* if @xmath6 is reducible , then , for any @xmath178 , @xmath3056 * if @xmath6 is @xmath107-antisymmetric , then @xmath3057 ; * if @xmath6 is reducible and @xmath107-antisymmetric , then @xmath3058 in the proof we will often use that @xmath3005 .
we begin with ( i ) .
it easily follows from lemma [ f ] that @xmath3059 for any @xmath28 and @xmath3060 .
hence , for any @xmath3061 , @xmath3062 therefore @xmath3063 similarly , for any @xmath3061 , @xmath3064 and it follows that @xmath3065 we have @xmath3066 where the signs @xmath3067 are computed by @xmath3068 therefore , using the cocommutativity of @xmath7 , we obtain @xmath3069 besides , we have @xmath3070 where the last equality uses the formula @xmath3071 which follows from the reducibility of @xmath6 .
we conclude that @xmath3072 .
we now prove ( ii ) .
if the bibracket @xmath6 is @xmath107-antisymmetric , then for any homogeneous @xmath39 , @xmath3073 where the last equality is a concequence of the cocommutativity of @xmath3055 and @xmath3074 finally , we deduce ( iii ) from ( ii ) : @xmath3075 we apply the definitions and results above in the geometric setting .
let @xmath0 be a smooth oriented connected manifold of dimension @xmath1921 with non - void boundary .
consider the path homology category @xmath2805 , see section [ phc ] , and assume that condition is satisfied .
the category @xmath2786 carries a natural structure of a graded hopf category .
the comultiplication @xmath534 in @xmath2786 is the direct sum over all @xmath2807 of the linear maps @xmath3076 induced by the diagonal map @xmath3077 .
the counit @xmath1328 in @xmath2786 is the augmentation defined in section [ scalar_form ] . for @xmath2133 , the inversion of paths induces a homeomorphism @xmath3078 which in its turn induces a graded linear isomorphism @xmath3079 ; the direct sum of these isomorphisms over all @xmath2133 defines an antipode @xmath1005 in @xmath2786 .
it is well - known that the path homology category @xmath3080 with this data is a cocommutative hopf category .
[ redred ] the intersection bibracket in @xmath2805 is reducible .
let @xmath2787 be the graded algebra associated with @xmath2786 and let @xmath3042 be the map associated with the intersection bibracket @xmath6 in @xmath2786 .
we must show that @xmath3081 for any @xmath178 . since @xmath3054 is bilinear , it suffices to consider the case where @xmath3082 and @xmath3083 for some @xmath1646 . here
@xmath3084 , and @xmath3085 are four points in @xmath3 .
observe that a path in @xmath92 starting from @xmath1949 represents a certain element @xmath3086 and by lemma [ f ] , @xmath3087 similarly , a path in @xmath92 ending at @xmath1948 represents a certain @xmath3088 and @xmath3089 thus it suffices to consider the case where @xmath2244 .
pick transversal smooth polycycles @xmath634 in @xmath1515 and @xmath712 in @xmath1516 representing respectively @xmath3090 and @xmath3091 .
we form the intersection polycycle @xmath3092 in @xmath1584 as in section [ def_d ] . by definition
, @xmath3093 corresponds to the homology class @xmath3094\in h_{p+q+2-n } \big(\omega_{32 } \times \omega_{14 } \big)$ ] under the isomorphism @xmath3095 induced by the cross product in homology .
consider the tensor @xmath3096 applying with @xmath1948 , @xmath2440 exchanged and with @xmath3097 , @xmath3098 , we obtain @xmath3099 where @xmath3100 is the isomorphism induced by the cross product in homology and @xmath3101 is induced by the diagonal map @xmath3077 .
the homology class @xmath3102 is represented by the polycycle @xmath3103 where @xmath3104 is obtained by projecting @xmath3105 onto @xmath1431 and applying @xmath3106 , whereas @xmath3107 is obtained by projecting onto @xmath3108 and applying @xmath3109 .
consider now the homeomorphisms @xmath3110 induced by the inversion of paths , the concatenation maps @xmath3111 , and the map @xmath3112 it follows that the image of the homology class @xmath3113 under the cross product isomorphism @xmath3114 is represented by the polycycle @xmath3115 . to analyze this polycycle ,
let @xmath3116 be the first and the second coordinates of @xmath3117 .
for any point @xmath3118 , the path @xmath3119 is obtained by concatenation of the following three paths : ( i ) the path @xmath3120 from @xmath1948 to @xmath3121 ; ( ii ) the initial segment of the path @xmath3122 from @xmath1949 to the point @xmath3123 ; ( iii ) the terminal segment of the path @xmath3124 from the latter point to @xmath2440 .
this concatenated path goes along a terminal segment of the path @xmath3120 twice in opposite directions .
therefore the path @xmath3125 is homotopic to a path @xmath3126 obtained by concatenation of just two paths : the initial segment of the path @xmath3120 from @xmath1948 to the point @xmath3123 and the terminal segment of the path @xmath3124 from the latter point to @xmath2440 .
the homotopy in question may be defined by an explicit formula which applies to all points @xmath3118 .
therefore , it determines a homotopy of the polycycle @xmath3127 into the polycycle @xmath3128 .
a similar argument applies to the path @xmath3129 and yields a homotopy of the polycycle @xmath3130 into @xmath3128 .
applying these two homotopies coordinatewise we obtain a homotopy of the polycycles @xmath3131 and @xmath3132 .
it is obvious that the homology class represented by the latter polycycle belongs to the image of the map @xmath3133 .
we conclude that @xmath3081 .
lemma [ redred ] and the @xmath2206-antisymmetry of the intersection bibracket of @xmath0 implies that it shares all the properties established in lemma [ fcoc ] .
note that the associated pairing @xmath74 generalizes the reidemeister pairing .
the results of this section are analogues of the known properties of the intersection bibracket in dimension two , see @xcite . in dimension two ,
the role of @xmath74 is played by the homotopy intersection form introduced in @xcite .
we show that a spherical boundary component of a manifold determines a moment map for the intersection bibracket .
this allows us to define an @xmath94-poisson structure on the pontryagin algebras of certain manifolds without boundary . for the notions of a moment map and an @xmath94-poisson structure we refer to appendix [ sectionactions- ] .
let @xmath0 be a smooth connected oriented manifold of dimension @xmath1921 with a fixed connected component @xmath982 of @xmath92 homeomorphic to the sphere @xmath2708 .
fix a point @xmath3134 and set @xmath3135 where @xmath3136 . the orientation - preserving homeomorphisms @xmath3137
represent an element @xmath3138 of @xmath3139 .
recall the connecting homomorphism @xmath3140 of section [ the pontryagin algebra ] and set @xmath3141 [ spherical_boundary ] if @xmath0 satisfies the condition , then @xmath3142 is a moment map for the intersection bibracket @xmath6 in @xmath2168 .
consider the path homology category @xmath2131 of @xmath0 and the intersection bibracket @xmath6 in the associated graded algebra @xmath3143 .
pick a smooth closed @xmath405-ball @xmath3144 and consider the smooth manifold @xmath3145 .
as above , we can consider the path homology category of @xmath3146 and the intersection bibracket @xmath3147 in the associated graded algebra .
consider the restriction of @xmath3148 to the algebra @xmath3149 the inclusion @xmath3150 induces a graded algebra homomorphism @xmath3151 .
the definition of the intersection bibracket implies that the following diagram commutes : @xmath3152^{{\left\{\!\!\left\{-,-\right\}\!\!\right\}}_{p } } \ar[d]_-{\iota \otimes \iota } & & b\otimes b \ar[d]^-{\iota \otimes \iota } \\ a \otimes a \ar[rr]^{{\left\{\!\!\left\{-,-\right\}\!\!\right\ } } } & & a \otimes a. } \ ] ] [ ] at 60 22 [ r ] at 75 65 [ ] at 58 117 [ ] at 519 129 [ b ] at 256 119 [ tl ] at 552 16 [ tr ] at 7 13 .,title="fig : " ] we must prove that @xmath3153 for any @xmath3154 . to this end , fix a path @xmath207 in @xmath3155 leading from @xmath60 to a distinct point @xmath3156 : see figure [ moment_map ] .
the path @xmath207 represents an element in @xmath3157 denoted also by @xmath207 .
this element is invertible , and its inverse @xmath3158 is represented by the inverse path .
set @xmath3159 ( in the notation of section [ change ] , @xmath3160 . ) clearly , @xmath3161 .
hence it suffices to prove that @xmath3162 the homology class @xmath3163 can be represented by a polycycle @xmath3164 in @xmath3165 . choosing the ball @xmath56 close enough to @xmath92
, we can ensure that it does not meet the image of @xmath3164 .
then @xmath3166 for the homology class @xmath3167 represented by @xmath3164 . similarly , @xmath3168 for some @xmath3169 .
we deduce from that @xmath3170 to proceed , we pick an embedded path in @xmath3146 leading from a point @xmath3171 to @xmath53 and meeting @xmath3172 only in the endpoints .
this path defines an invertible element @xmath3173 .
set @xmath3174 note that @xmath3175 since @xmath3176 can be represented by a polycycle whose image does not meet @xmath982
. therefore @xmath3177 by lemma [ intersecting_arc_and_spheres ] , we obtain @xmath3178 .
therefore @xmath3179 .
we conclude that @xmath3180 which proves .
note the following consequence of lemma [ spherical_boundary ] : under the assumptions of this lemma , the following three statements are equivalent : * @xmath3181 ; * @xmath3182 ; * the intersection bibracket in @xmath2205 is zero .
let @xmath2718 be a smooth connected oriented manifold of dimension @xmath1921 without boundary .
our construction of a bibracket in the pontryagin algebra of a manifold requires the base point to lie in the boundary and does not apply to @xmath2718 .
one way to circumvent this difficulty is to cut out a small open ball from @xmath2718 . if the resulting manifold satisfies then its pontryagin algebra carries the intersection bibracket .
this algebra with bibracket is a diffeomorphism invariant of @xmath2718 . under certain assumptions on @xmath2718
we can use the hamiltonian reduction of appendix [ sectionactions- ] to define a @xmath94-poisson structure on the pontryagin algebra of @xmath2718 . to this end , pick a base point @xmath3183 and a smooth closed @xmath405-ball @xmath3184 with @xmath3185 . consider the smooth manifold @xmath3186 with @xmath3187 .
let @xmath3188 be the pontryagin algebras of @xmath3189 and @xmath2718 , respectively .
the inclusion @xmath3190 induces a graded algebra homomorphism @xmath3191 , clearly , @xmath3192 where @xmath3193 .
therefore , @xmath3194 .
[ no_boundary ] assume that @xmath3186 satisfies the condition , the homomorphism @xmath3195 is onto , and @xmath3196 .
then the intersection bibracket @xmath6 in @xmath2168 induces an @xmath94-poisson structure of degree @xmath1674 on @xmath3197 .
this structure does not depend on the choice of the ball @xmath56 .
the first claim follows from lemma [ quotient ] .
the independence of the choice of the ball is a consequence of the naturality of the intersection bibracket under diffeomorphisms , and the fact that for any balls @xmath3198 with @xmath3199 there is a diffeomorphism @xmath3200 such that @xmath3201 , @xmath3202 , and @xmath259 is isotopic to @xmath3203 in the class of diffeomorphisms @xmath3204 fixing @xmath60 .
such an @xmath259 acts on @xmath3197 as the identity , and the result follows .
note that an @xmath94-poisson structure on @xmath3197 induces gerstenhaber brackets on the trace algebras of @xmath3197 , see appendix [ sectionactions- ] .
there are manifolds which do not satisfy the assumptions of theorem [ no_boundary ] , for example , @xmath3205 . nonetheless , according to @xcite and @xcite , these assumptions are satisfied
if @xmath2718 is a closed simply connected manifold whose cohomology algebra @xmath3206 is not generated by a single element and @xmath95 is a field whose characteristic is equal to zero or is sufficiently big .
we consider the example of section [ sc_example ] and keep the same notation .
thus @xmath3207 and @xmath3208 .
the element @xmath3209 can be computed as follows . as a topological manifold , @xmath0 is the boundary - connected sum of the manifolds @xmath3210 where @xmath3211 . hence @xmath3212 where @xmath3213 is the inclusion homomorphism ( we can ignore the base point because @xmath3214 and @xmath0 are simply - connected ) . by the definition of the whitehead bracket @xmath2215 _ { \operatorname{wh } }
$ ] in @xmath2216 , we have @xmath3215 _ { \operatorname{wh } } $ ] where @xmath3216 and @xmath3217 are represented by the two factors of @xmath3214 .
thus , @xmath3218 _ { \operatorname{wh } } + \cdots + [ x_g^\pi , y_g^\pi ] _ { \operatorname{wh } } \in \pi_{n-1}(m).\ ] ] recall that the bracket @xmath2215 $ ] in @xmath2168 induced by the pontryagin multiplication is related to the whitehead bracket in @xmath2216 by the formula @xmath3219
= \left[\ , \overline \partial_{p_i}(x_i^\pi ) , \overline \partial_{q_i}(y_i^\pi ) \right ] = ( -1)^{p_i}\ , \overline \partial_{p_i+q_i-1 } \big ( [ x_i^\pi , y_i^\pi ] _ { \operatorname{wh } } \big ) \in a_\star.\ ] ] therefore @xmath3220 + \cdots + ( -1)^{p_g } [ x_g , y_g ] \in a_\star.\ ] ] a direct computation on the generators @xmath2737 of @xmath2168 using confirms that @xmath3117 is a moment map of the intersection bibracket @xmath6 of @xmath0 , as claimed by lemma [ spherical_boundary ] .
consider in more detail the case @xmath3221 and set @xmath3222 the loop space of @xmath3223 based at @xmath49 is the product of the loop spaces of @xmath2627 and @xmath2631 . by the knneth theorem , the pontryagin algebra @xmath3224 is ( as a graded algebra ) the tensor product of the pontryagin algebras of @xmath2627 and @xmath2631 .
since the graded algebra @xmath2168 is freely generated by @xmath2774 , the quotient @xmath3225 is the commutative graded algebra freely generated by @xmath2774 .
it is clear that the assumptions of theorem [ no_boundary ] are satisfied here for any ground ring @xmath95 .
theorem [ no_boundary ] yields an @xmath94-poisson structure @xmath2207 of degree @xmath1674 on the ( commutative ) graded algebra @xmath3197 .
the bracket @xmath2207 in @xmath3226 is then a gerstenhaber bracket of degree @xmath1674 .
it coincides with the gerstenhaber bracket @xmath38 in @xmath3227 computed in section [ sc_example ] .
we consider the example of section [ nsc_example ] and keep the same notation .
thus , @xmath2750 and @xmath3228 .
the element @xmath3209 can be computed as follows .
consider the cylinder @xmath3229 with the product orientation and pick a closed @xmath405-ball @xmath2717 in its interior .
then @xmath3230 = [ \{1\ } \times s^{n-1 } ] - [ \{0\ } \times s^{n-1 } ] \in \pi_{n-1}\big ( ( i \times s^{n-1 } ) \setminus { \operatorname{int}}({d})\big).\ ] ] ( here @xmath56 carries the orientation induced by @xmath0 and @xmath3231 carries the orientation inherited from @xmath56 . ) it follows that @xmath3232 where the dot denotes the action of @xmath2183 on @xmath2780 .
we deduce that @xmath3233 a direct computation using confirms that @xmath3117 is a moment map of the intersection bibracket @xmath6 of @xmath0 , as claimed by lemma [ spherical_boundary ] . by the k " unneth theorem ,
the pontryagin algebra @xmath3197 of @xmath93 is ( as a graded algebra ) the tensor product of the pontryagin algebras of @xmath3234 and @xmath3235 .
thus , @xmath3197 is the commutative graded algebra freely generated by @xmath3236 and @xmath3237 .
as a consequence , the assumptions of theorem [ no_boundary ] are satisfied ( for any ground ring @xmath95 ) so that the intersection bibracket @xmath6 of @xmath0 induces an @xmath94-poisson structure @xmath2742 of degree @xmath1674 on @xmath3197 . since @xmath3197 is commutative , this structure is a gerstenhaber bracket of degree @xmath1674 . according to
, it is given by @xmath3238 the results of this section are high - dimensional analogues of the well - known properties of surfaces .
the pontryagin algebra of a closed connected oriented surface is the group algebra @xmath3239 $ ] where @xmath2238 is the fundamental group of the surface .
then @xmath3240= { \mathbb{k } } [ \check \pi]$ ] is the module freely generated by the set @xmath3241 of conjugacy classes in @xmath11 .
the goldman lie bracket in @xmath3242 is the canonical @xmath94-poisson structure on @xmath24 .
we discuss certain brackets induced by bibrackets and relate them to the trace map . in the non - graded setting , these results are due to van den bergh @xcite . in this appendix , @xmath7 is a graded algebra .
any bibracket @xmath6 in @xmath7 induces a bracket @xmath3243 called the _ associated bracket _ by @xmath3244 [ comparis- ]
let @xmath6 be a @xmath107-antisymmetric @xmath107-graded bibracket in @xmath7
. then the associated bracket @xmath3245 has the following properties : * @xmath2742 has degree @xmath107 and satisfies the @xmath107-graded leibniz rule , * @xmath3246)$ ] for all homogeneous @xmath39 , * @xmath3247 , a \rangle=0 $ ] and @xmath3248\rangle \subset [ a , a]$ ] , * for any homogeneous @xmath198 , @xmath3249 where @xmath3250 carries @xmath3251 to @xmath3252 for all @xmath3253 . claim ( i ) is straightforward . to check ( ii ) , set @xmath3254
. then @xmath3255 by and , modulo @xmath123 $ ] , @xmath3256 to check ( iii ) , pick any homogeneous elements @xmath198 and set @xmath3257 , @xmath3258 .
we have @xmath3259 so that @xmath3260 transposing @xmath32 and @xmath122 , we also obtain @xmath3261 since @xmath3262 has degree @xmath107 , we have @xmath3263 and @xmath3264 . therefore @xmath3265
. hence @xmath3247 , a \rangle=0 $ ] .
this equality together with ( ii ) imply the inclusion @xmath3248\rangle \subset [ a , a]$ ] .
we now prove ( iv ) .
set @xmath3266 , @xmath3267 , @xmath3268 , @xmath3269 , and @xmath3270 .
then @xmath3271 we deduce that @xmath3272 by ( i ) , we have @xmath3273 and @xmath3274 by the definition of the tribracket , @xmath3275 therefore @xmath3276 transposing @xmath3277 , we obtain @xmath3278 then ( iv ) follows from .
lemma [ comparis- ] implies that the bracket @xmath3279 in @xmath7 associated with @xmath6 induces a @xmath107-antisymmetric bracket of degree @xmath107 in the graded module @xmath3280 $ ] .
the latter bracket is again denoted by @xmath3279 . for any @xmath16 , there is a canonical linear map @xmath3281 defined by @xmath3282 for @xmath28 . clearly , @xmath3283)=0 $ ] so that @xmath3284 induces a linear map from @xmath3285 $ ] to @xmath3286 .
the latter map is also denoted by @xmath3284 and called the _
trace_. the graded subalgebra of @xmath35 generated by @xmath3287 is denoted by @xmath3288 and called the _
@xmath34-th trace algebra _ of @xmath7 .
[ comparis ] under the conditions of lemma [ comparis- ] , the map @xmath3289 carries the bracket @xmath3290 in @xmath3291 associated with @xmath6 into the bracket @xmath38 in @xmath35 induced by @xmath6 .
pick any @xmath39 and let @xmath3292 be their projections to @xmath3291 .
we have @xmath3293 lemma [ comparis ] implies that @xmath3294 for all @xmath16 . for @xmath147 ,
the trace map @xmath3295 is the canonical projection and @xmath3296 .
if @xmath7 is unital and @xmath3297 is the unital graded subalgebra of @xmath316 generated by the projection of @xmath3298 to @xmath316 , then similarly @xmath3299 .
we formulate hamiltonian reduction for gerstenhaber bibrackets based on a notion of an @xmath94-poisson structure on a graded algebra . in the non - graded case ,
the results of this appendix are due to crawley - boevey @xcite and van den bergh @xcite .
an _ @xmath94-poisson structure of degree @xmath106 _ on a graded algebra @xmath7 is a @xmath107-graded lie bracket @xmath3300 in the graded module @xmath3280 $ ] such that , for all homogeneous @xmath3301 , the map @xmath3302 lifts to a derivation @xmath3007 of degree @xmath3303 . if @xmath7 is a commutative graded algebra , then an @xmath94-poisson structure of degree @xmath3304 in @xmath7 is nothing but a gerstenhaber bracket of degree @xmath107 in @xmath7 .
[ vdb_to_cb ] given a gerstenhaber bibracket of degree @xmath107 in a graded algebra @xmath7 , the induced bracket @xmath3305 in @xmath3306 is an @xmath94-poisson structure of degree @xmath107 on @xmath7 . that @xmath3279 is a @xmath107-graded lie bracket in @xmath3291 follows from lemma [ comparis- ] .
the same lemma shows that the formula @xmath3307 defines a linear map @xmath3308 which preserves the lie bracket and carries @xmath3309 to @xmath3310 for all @xmath152 .
this implies the claim of the lemma .
the following theorem is due to crawley - boevey in the non - graded case .
[ h0_to_g ] let @xmath3300 be an @xmath94-poisson structure of degree @xmath3304 on a graded algebra @xmath7 and let @xmath16 .
then there is a unique gerstenhaber bracket @xmath38 of degree @xmath107 in the trace algebra @xmath3311 such that @xmath3312 for any @xmath3313 .
the proof follows the same lines as in the non - graded case , see ( * ? ? ?
* theorem 4.5 ) .
the uniqueness of @xmath38 is obvious because the image of the trace map @xmath3289 generates @xmath3314 . to prove the existence ,
consider the commutative graded algebra @xmath3315 freely generated by the graded module @xmath3291 ( the _ symmetric algebra _ of @xmath3291 ) .
the @xmath107-graded lie bracket @xmath2207 in @xmath3291 uniquely extends to a gerstenhaber bracket @xmath3316 of degree @xmath107 in @xmath1117 .
the map @xmath3317 uniquely extends to a graded algebra homomorphism @xmath3318 , which is surjective .
therefore , it suffices to prove the existence of a map @xmath3319 such that the following diagram commutes : @xmath3320 ^ -{\langle-,-\rangle_s } \ar[d]_-{t \times t } & & s \ar[d]^-{t } \\
a_n^t \times a_n^t \ar[rr]^-{{\left\{-,-\right\ } } } & & a_n^t\ , . }
\ ] ] in other words , we need to show that the pairing @xmath3321 annihilates @xmath3322 .
since the bracket @xmath3316 is @xmath107-antisymmetric , it suffices to show that @xmath3323 for any @xmath3324 .
since the bracket @xmath3316 satisfies the @xmath107-graded leibniz rule in the first variable and @xmath3325 is an algebra homomorphism , it suffices to consider the case @xmath3326 . by the definition of an @xmath94-poisson structure , the map @xmath3327 lifts to a derivation @xmath3328 .
there is a unique derivation @xmath3329 such that @xmath3330 for any @xmath28 and @xmath3331 .
then , for any @xmath28 , @xmath3332 it follows that the maps @xmath3333 and @xmath3334 are equal on @xmath3335 .
since @xmath3291 generates the algebra @xmath982 and both these maps are derivations , they must be equal . as a consequence , @xmath3323 .
combining lemma [ vdb_to_cb ] and theorem [ h0_to_g ] , we obtain that any gerstenhaber bibracket of degree @xmath107 in @xmath7 induces a gerstenhaber bracket of degree @xmath107 in @xmath3288 .
clearly this bracket is the restriction of the gerstenhaber bracket in @xmath35 provided by lemma [ importantgers ] , cf .
section [ hamham1 ] .
let @xmath7 be a unital graded algebra equipped with a gerstenhaber bibracket @xmath6 of degree @xmath107 .
a _ moment map _ for @xmath6 is an element @xmath3336 such that @xmath3337 for all @xmath28 or , equivalently , @xmath3338 for all @xmath28 .
if @xmath3339 , then there is at most one moment map . if @xmath3340 , then for any moment map @xmath3341 and any @xmath138 , the sum @xmath3342 is a moment map . in the non - graded setting , the following lemma is due to van den bergh @xcite .
[ quotient ] let @xmath3336 be a moment map .
the bracket @xmath2742 in @xmath3343 associated with @xmath6 induces an @xmath94-poisson structure of degree @xmath107 on @xmath3344 .
let @xmath3345 and @xmath3346 $ ] be the canonical projections .
clearly , @xmath102 carries @xmath123 $ ] to @xmath3347 $ ] and induces a linear map @xmath3348 .
lemma [ comparis-](iii ) shows that the bracket @xmath3349 in @xmath7 induces a pairing @xmath3350 .
we claim that there are linear maps @xmath3351 such that the following diagram commutes : @xmath3352 ^ - { { \operatorname{id}}\otimes p } \ar[d]_-{\langle -,-\rangle } & & \check{a } \otimes { b } \ar[d]_-{u } \ar[rr]^-{\check p \otimes h } & & \check{{b } } \otimes \check { b } \ar[d]^-{v } \\ a \ar[rr]^- { p } & & { b } \ar[rr]^- { h } & & \check { b}\ , . } \ ] ] such maps @xmath3353 are necessarily unique because @xmath1429 , @xmath3354 , @xmath3355 are onto . as a consequence , the following diagram commutes : @xmath3356_-{\langle -,-\rangle } \ar[rr]^-{\check p \otimes \check p } & & \check{{b } } \otimes \check{{b } } \ar[d]^-{v } \\ \check a \ar[rr]^- { \check p } & & \check{{b}}\ , . } \ ] ] therefore @xmath1535 is a @xmath107-graded lie bracket in @xmath3242 .
since @xmath3357 is a derivation for all @xmath3358 and @xmath3354 is onto , the lie bracket @xmath1535 is an @xmath94-poisson structure on @xmath3359 .
it remains to verify the claim above .
the definitions of the moment map @xmath3117 and the bracket @xmath2207 in @xmath115 imply that @xmath3360 . hence , @xmath3361 .
this inclusion implies the existence of @xmath797 . by lemma [ comparis-](ii ) , @xmath3362={\operatorname{ker}}h p.\ ] ] this implies the existence of @xmath1535 . by theorem [ h0_to_g ]
, we obtain that under the assumptions of lemma [ quotient ] , the bibracket in @xmath7 induces gerstenhaber brackets of degree @xmath107 on the trace algebras of @xmath3363 . as an exercise ,
the reader may extend lemma [ quotient ] to the setting of graded categories discussed in section [ extenscatmain ] .
we discuss the behavior of the bracket constructed in lemma [ fdbtopb ] under the natural actions of the general linear group and the lie algebra of matrices .
we begin with terminology . by a _ lie pair _ we mean a pair @xmath3364 where @xmath12 is a group and @xmath3365 is a ( non - graded ) lie algebra endowed with a ( left ) action of @xmath12 on @xmath3365 by lie algebra automorphisms .
the action is denoted by @xmath3366 for @xmath3367 and @xmath3368 . given a lie pair @xmath3369 , by a _
@xmath3369-algebra _ we mean a graded algebra @xmath7 endowed with an action of @xmath3370 and an action of @xmath3365 such that @xmath3371 for all @xmath3372 , @xmath3367 , @xmath28 . here
an action of @xmath12 on @xmath7 is a group homomorphism from @xmath12 to the group of graded algebra automorphisms of @xmath7 , and an action of @xmath3365 on @xmath7 is a lie algebra homomorphism from @xmath3365 to the lie algebra of derivations of @xmath7 of degree zero , cf@xmath134 section [ gral++++ ] .
fix an integer @xmath16 .
let @xmath3373 be the @xmath34-th general linear group over @xmath95 and let @xmath3374 be the lie algebra of @xmath151-matrices with lie bracket @xmath3375=uv - vu$ ] .
the group @xmath3376 acts on @xmath3377 by @xmath3378 for any @xmath3379 , @xmath3380 , which turns the pair @xmath3381 into a lie pair .
we explain now that the graded algebra @xmath135 associated with an arbitrary graded algebra @xmath7 in section [ amt0 ] has a canonical structure of a @xmath3382-algebra .
the group @xmath3383 acts on @xmath135 as follows : for a matrix @xmath3384 and a generator @xmath3385 , set @xmath3386 in this formula , the numerical coefficients appear to the left of the generator @xmath3387 .
it is easier to remember in the equivalent form @xmath3388 , and we will use the latter form .
direct computations show that these formulas are compatible with the relations in @xmath135 and define an action of @xmath3376 on @xmath135 .
we verify the compatibility with the relation @xmath3389 : @xmath3390 the lie algebra @xmath3391 acts on @xmath135 as follows : for a matrix @xmath3392 and a generator @xmath3393 , set @xmath3394 this formula is compatible with the relations in @xmath135 and defines an action of @xmath3377 on @xmath135 .
we verify the compatibility with the relation @xmath3389 : @xmath3395 it is easy to check that these actions turn @xmath135 into a @xmath3382-algebra .
moreover , these actions descend to the commutative graded algebra @xmath3396 and turn it into a @xmath3381-algebra .
if @xmath7 is unital , then similar claims hold for the unital graded algebras @xmath3397 and @xmath3398 . in the non - graded setting , the following lemma is due to van den bergh @xcite .
[ fdbtopb+action ] let @xmath6 be a @xmath107-graded bibracket in a graded algebra @xmath7 .
for any @xmath16 , the bracket @xmath200 in @xmath3399 defined in lemma [ fdbtopb ] satisfies @xmath3400 for all @xmath3401 , @xmath3402 and @xmath3403 .
pick @xmath3404 .
it is easy to see that if the identity @xmath3405 holds for all the generators of @xmath3286 , then it holds for any @xmath3406 . given @xmath39 and @xmath3407 , @xmath3408 similarly , given @xmath3409 , it is enough to check the identity @xmath3410 for the generators of @xmath35 . for @xmath39 and @xmath3407 , @xmath3411 similar results
hold in the setting of graded categories discussed in section [ extenscatmain ] ; we leave the details to the reader . for any graded algebra @xmath7 , we have @xmath3412 where @xmath3413 is the trace algebra and @xmath3414 is the subalgebra of @xmath35 consisting of all elements invariant under the action of @xmath3376 .
when @xmath7 is finitely generated as an algebra and the ground ring @xmath95 is a field of characteristic zero , @xmath3415 , see @xcite .
in this appendix , we relate the intersection bibracket to the chas
sullivan string bracket in loop homology of a manifold .
let @xmath0 be a smooth oriented connected manifold of dimension @xmath1509 with non - empty boundary . by a _ loop _ in @xmath0 we mean a continuous map @xmath3416 where @xmath3417
let @xmath3418 be the space of loops in @xmath0 with compact - open topology .
the _ loop homology _ of @xmath0 is @xmath3419 .
the _ string homology _ , @xmath3420 , of @xmath0 is the @xmath3421-equivariant homology of @xmath3422 where @xmath3421 acts on @xmath3423 by @xmath3424 for any @xmath3425 and @xmath3426 .
thus , @xmath3427 where @xmath3428 is the total space of the universal @xmath3421-principal fiber bundle and @xmath3429 is the quotient of @xmath3430 by the diagonal action of @xmath3421 . since @xmath3431 is contractible , the projection @xmath3432 induces a linear map @xmath3433 .
chas and sullivan @xcite defined a degree @xmath1674 lie bracket in @xmath3420 called the _
string bracket_. for @xmath1970 , this is the goldman bracket discussed in section [ dimension_2 ] .
we assume that @xmath2749 and relate the string bracket to the lie bracket @xmath2207 in @xmath3434 $ ] defined in section [ the induced lie bracket ] .
[ maprr ] let @xmath47 , @xmath3435 , and let @xmath3436 be the inclusion map .
the induced homology homomorphism @xmath3437 annihilates @xmath2202 $ ] and induces a linear map @xmath3438 .
the composition @xmath3439 is a graded lie algebra homomorphism .
let @xmath3440 be the concatenation of loops . for @xmath3441 , @xmath3442 , @xmath3443 where @xmath3444 is the transposition .
therefore , to show that @xmath3445 , it suffices to prove that @xmath3446 . clearly , @xmath3447 where @xmath3448 stands for the action of @xmath3449 . since @xmath3450 is homotopic to the identity , @xmath3451 is homotopic to @xmath3452 .
we deduce that @xmath3453\right)=0 $ ] .
recall the definition of the string bracket @xmath3454 $ ] in @xmath3420 .
let @xmath3455 be the degree 1 lift map in the gysin sequence of the @xmath3421-bundle @xmath3456 .
chas and sullivan define a linear map @xmath3457 of degree @xmath3458 called the _ loop product _ ( and denoted by @xmath2810 in @xcite ) . for a detailed exposition , the reader is referred for instance to cieliebak @xcite . for a homogoneous @xmath3459 and any @xmath3460 , @xmath3461= ( -1)^{\vert x \vert+n } \operatorname{e } \big(\operatorname{m}(x ) \bullet _ { \operatorname{cs } } \operatorname{m}(y)\big ) .\ ] ] this formula implies that to prove the second claim of the lemma , it is enough to show the commutativity of the diagram @xmath3462_-{(-1)^{\vert-\vert}\operatorname{m } \ !
\operatorname{e}\ ! \operatorname{r
} \otimes \operatorname{m } \ ! \operatorname{e}\ !
\operatorname{r } } \ar[r]^-{\langle -,-\rangle } & \check a_\star \ar[d]^-{\operatorname{r}}\\ { \mathbb{h}}\otimes { \mathbb{h}}\ar[r]^- { \bullet _ { \operatorname{cs } } } & { \mathbb{h}}. } \ ] ]
let @xmath3463 and @xmath3464 with @xmath1646 .
let @xmath3465 be the canonical projection . to compute @xmath3466
, we pick a path @xmath61 in @xmath92 connecting @xmath60 to another point @xmath63 . by definition , @xmath3467\big ) \\ & = & ( -1)^{q + n p } h { \mathsf{c } } _ * \big(\big [ \left((\varsigma^{-1},e_\star)_\sharp \times ( e_\star,\varsigma^{-1})_\sharp\right ) \widetilde\upsilon({\left\langle}a { \right\rangle}\otimes ( \varsigma,\varsigma)_\sharp { \left\langle}b { \right\rangle})\big]\big).\end{aligned}\ ] ] let @xmath634 be a reduced smooth polycycle in @xmath3468 and let @xmath712 be a reduced smooth polycycle in @xmath3469 such that @xmath2266 transversely represents the pair @xmath3470 . consider the intersection polychain @xmath2304 . then @xmath3471 \\ \notag & = & ( -1)^{q + n p } \big [ r _ * { \mathsf{c } } _ * \left ( ( \varsigma^{-1},e_\star)_\sharp \times ( e_\star,\varsigma^{-1})_\sharp \right ) { \mathscr{d}}({\mathscr{k}},{\mathscr{l}})\big ] \\ \notag & = & ( -1)^{q +
n p } \big[\big(d,\theta , w , r { \mathsf{c}}\left ( ( \varsigma^{-1},e_\star)_\sharp \times ( e_\star,\varsigma^{-1})_\sharp \right)(\kappa { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda)\big ) \big ] \\
\notag & = & ( -1)^{q +
n p } \big[\big(d,\theta , w , r ( \varsigma^{-1 } , \varsigma^{-1})_\sharp { \mathsf{c}}(\kappa { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda)\big ) \big ] \\
\notag & = & ( -1)^{q +
n p } \big[\big(d,\theta , w , r ' { \mathsf{c}}(\kappa { \!\mathrel{{\triangleleft}{\triangleright}}\!}\lambda)\big ) \big]\end{aligned}\ ] ] where , in the last two lines , @xmath3472 is the concatenation of paths @xmath3473 and @xmath3474 is the inclusion . on the other hand , @xmath3475 = \operatorname{me } \left[(k,\varphi , u , r\kappa)\right].\ ] ] using the computation of the map @xmath3476 in @xcite ( where this map is denoted by @xmath534 ) , we obtain @xmath3477\ ] ] where we use the following notation : @xmath3478 is the partition on @xmath3479 induced by @xmath443 ( by identifying @xmath3480 to @xmath3481 via @xmath3482 for any faces @xmath449 of the same type in @xmath409 ) ; @xmath3483 is the weight on @xmath3479 induced by @xmath797 via the equality @xmath3484 ; the map @xmath3485 is defined using the action of @xmath3421 on @xmath3422 by @xmath3486 for @xmath3487 and @xmath3488 .
the sign @xmath740 in is caused by a permutation of the two factors of @xmath3479 with respect to @xcite .
similarly , @xmath3489
= \operatorname{me } \left [ r'_*\ ! ( \varsigma,\varsigma)_\sharp { \left\langle}b { \right\rangle}\right ] = ( -1)^q \left[(l \times s^1 , \bar \psi , \bar v , \bar \lambda ) \right]\ ] ] where the map @xmath3490 is defined by @xmath3491 for @xmath3492 and @xmath3488 . the loop product @xmath3493 can be computed in terms of face homology .
this gives @xmath3494 \bullet_{\operatorname{cs } } \left[(l \times s^1 , \bar \psi , \bar v , \bar \lambda ) \right]\\ \notag & = & ( -1)^ { p+q + np } \left[(\bar d , \bar \theta , \bar w , \bar \kappa \infty \bar \lambda ) \right ] .\end{aligned}\ ] ] here @xmath3495 is the inverse image of @xmath1317 under the map @xmath3496 note that @xmath3495 has a structure of a manifold with faces inherited from @xmath3497 .
the orientation , the partition @xmath3498 and the weight @xmath3499 of @xmath3495 are as in the definition of the intersection operation @xmath1591 in section [ def_d ] .
the map @xmath3500 sends a point @xmath3501 to the loop that first goes along the loop @xmath3502 and then along the loop @xmath3503 .
note the sign @xmath3504 in , which arises from the difference between our orientation conventions and those in @xcite .
the map @xmath3505 determined by the canonical projection @xmath3506 induces an orientation - preserving diffeomorphism @xmath3507 which carries the partition @xmath1580 into @xmath3498 and the weight @xmath1694 into @xmath3499 . using the action of @xmath3508 on @xmath3422 , one
easily constructs a homotopy between the maps @xmath3509 and @xmath3510 from @xmath3511 to @xmath3423 .
it follows that @xmath3512 = \left[(\bar d , \bar \theta , \bar w , \bar \kappa \infty \bar \lambda ) \right ] \in { \mathbb{h}},\ ] ] and we deduce from and .
y. berest , x. chen , f. eshmatov , a. ramadoss , _ noncommutative poisson structures , derived representation schemes and calabi yau algebras .
_ mathematical aspects of quantization , 219246 , contemp .
, 583 , amer .
soc . , providence , ri , 2012 .
v. v. fock , a. a. rosly , _ poisson structure on moduli of flat connections on riemann surfaces and the r - matrix .
_ ( russian ) moscow seminar in math .
english translation : amer .
transl . ser . 2 , 191 , 6786 ( 1999 ) . | given a smooth oriented manifold @xmath0 with non - empty boundary , we study the pontryagin algebra @xmath1 where @xmath2 is the space of loops in @xmath0 based at a distinguished point of @xmath3 . using the ideas of string topology of chas
sullivan , we define a linear map @xmath4 which is a double bracket in the sense of van den bergh satisfying a version of the jacobi identity . for @xmath5 ,
the double bracket @xmath6 induces gerstenhaber brackets in the representation algebras associated with @xmath7 .
this extends our previous work on the case @xmath8 where @xmath9 is the group algebra of @xmath10 and @xmath6 induces the classical poisson brackets on the moduli spaces of representations of @xmath10 . |
for @xmath0 a closed surface and @xmath1 two regularly homotopic generic immersions , we are interested in the number mod 2 of quadruple points occurring in generic regular homotopies between @xmath2 and @xmath3 .
it has been shown in [ n ] that this number is the same for all such regular homotopies , and so it is a function of @xmath2 and @xmath3 which we denote @xmath4 . there
then arises the problem of finding explicit formulae for @xmath14 .
assuming @xmath0 is orientable , we give an explicit formula for @xmath15 , where @xmath5 is any generic immersion and @xmath6 is any diffeomorphism such that @xmath2 and @xmath7 are regularly homotopic ( theorem main1 ) .
for two special cases a formula for @xmath15 has already been known : the case where @xmath0 is a sphere has appeared in [ mb ] and [ n ] , and the case where @xmath0 is a torus and @xmath2 is an embedding has appeared in [ n ] . based on the smale - hirsch theorem , pinkall in [ p ] gave a useful tool for determining when two immersions are regularly homotopic , namely , any immersion @xmath5 induces a quadratic form @xmath16 , and two immersions @xmath1 are regularly homotopic iff @xmath17 .
let @xmath18 denote the group of all diffeomorphisms @xmath6 up to isotopy .
given @xmath5 we are interested in the group of all @xmath19 such that @xmath15 is defined , that is the group of all @xmath19 such that @xmath2 and @xmath7 are regularly homotopic .
it follows from the above criterion that this is precisely the group @xmath20 of all @xmath19 which preserve the quadratic form @xmath21 on @xmath11 .
we are thus lead to study the groups @xmath22 , starting with their index 2 subgroup @xmath23 of orientation preserving maps .
the plan of the paper is as follows : in section a we present the known results on quadratic forms which we will need . in section b
we show that the expression @xmath24 appearing in our proposed formula for @xmath15 defines a homomorphism on appropriate subgroups of @xmath25 ( theorem t1 ) . in section c
we show that ( except for one special case ) the group @xmath23 is generated by dehn twists and squares of dehn twists ( theorem t3 ) .
our main result is then proved in section d. we start with surfaces of genus 0 and 1 where we compute @xmath15 by giving explicit regular homotopies for generators of @xmath20 .
we then continue by induction on the genus , using the special nature of the generators of @xmath26 that we have found and of an additional generator for @xmath20 .
each generator respects a separation of @xmath0 into surfaces of smaller genus , a fact which will enable us to construct regular homotopies for @xmath0 by combining regular homotopies of surfaces of smaller genus .
in this section we summarize the definitions and known properties of quadratic forms over @xmath27 which will be needed in our work .
proofs to all facts stated in this section may be found in [ c ] , except for those relating to the arf invariant , which may be found in [ l ] .
let @xmath28 be a finite dimensional vector space over @xmath27 .
a function @xmath29 is called a _
quadratic form _
if @xmath30 satisfies : @xmath31 for all @xmath32 , where @xmath33 is a bilinear form .
the following properties follow : ( a ) @xmath34 .
( b ) @xmath35 for all @xmath36 .
( c ) @xmath37 for all @xmath38 .
@xmath30 is called _ non - degenerate _
if @xmath39 is non - degenerate , i.e. for any @xmath40 there is @xmath41 with @xmath42 .
[ pc1 ] if @xmath30 is non - degenerate then @xmath28 is necessarily of even dimension and there exists a basis @xmath43 for @xmath28 such that @xmath44 and @xmath45 for all @xmath46 and such that one of the following two possibilities holds : 1 .
@xmath47 for @xmath48 .
2 . @xmath49 and @xmath47 for @xmath50 .
@xmath30 is completely determined by the values @xmath51 and @xmath52 on a basis @xmath53 and so for given dimension @xmath54 there are two isomorphism classes of non - degenerate quadratic forms , and they are in fact distinct .
the invariant @xmath55 is then defined to be 0 or 1 according to whether 1 or 2 of proposition pc1 holds respectively .
( in the more general setting of [ c ] , this is equivalent to @xmath30 having index @xmath12 or @xmath56 respectively . ) the arf invariant is additive in the following sense : [ pc2 ] if @xmath57 , @xmath58 , are non - degenerate quadratic forms , then @xmath59 defined by @xmath60 is a non - degenerate quadratic form with @xmath61 .
from now on we will always assume that our quadratic form @xmath30 is non - degenerate .
[ pc3 ] if @xmath62 are independent and @xmath63 for all @xmath64 then there are @xmath65 with @xmath66 and @xmath45 for all @xmath64 .
( @xmath67 are then necessarily independent . )
a linear map @xmath68 is called _ orthogonal _ with respect to @xmath30 if @xmath69 for all @xmath36 .
it then follows that @xmath70 for all @xmath38 and that @xmath71 is invertible .
the group of all orthogonal maps of @xmath28 with respect to @xmath30 will be denoted @xmath72 .
[ d0 ] given @xmath73 , define @xmath74 by @xmath75 . [ pc4 ] @xmath76 iff @xmath77 or @xmath78 .
[ tc ] except for the case when @xmath79 and @xmath80 , @xmath72 is generated by the elements @xmath81 with @xmath77 .
theorem tc will also follow from theorem t3 below .
see remark r5 .
if @xmath82 is a subspace , then the conjugate space @xmath83 of @xmath84 is defined by @xmath85 . if @xmath73 we similarly define @xmath86 .
let @xmath87 denote the identity map on @xmath28 , let @xmath88 denote the image of @xmath71 and let @xmath89 .
[ pc5 ] if @xmath90 then @xmath91 .
let @xmath73 , then @xmath92 and so if @xmath93 @xmath94 , where @xmath95 .
[ p1 ] let @xmath96 and @xmath73 with @xmath77 . 1 .
if @xmath97 then @xmath98 .
2 . if @xmath99 then @xmath100 .
we first note : ( a ) if @xmath101 then @xmath102 so @xmath103 and so @xmath104 .
( b ) for any @xmath36 , if @xmath105 then @xmath106 and so @xmath107 , that is @xmath101 .
we get that @xmath108 iff @xmath109 with @xmath105 iff @xmath110 with @xmath105 iff @xmath110 with @xmath111 iff @xmath112 ( proposition pc5 ) iff @xmath113 . since @xmath114
we conclude that @xmath115 iff @xmath97 .
clearly @xmath116 and let @xmath117 denote the dimension of this subspace .
@xmath118 is of codimension 1 in @xmath28 and so it follows : ( 1 ) if @xmath97 then @xmath119 and @xmath115 and so @xmath120 . ( 2 ) if @xmath99 then @xmath121 and @xmath122 and so @xmath123 .
we now define @xmath124 by : @xmath125 [ r ] since @xmath126 ( or by proposition pc5 ) we may also write : @xmath127 , and since @xmath28 is of even dimension we also have : @xmath128 . [ t1 ]
@xmath124 is a ( non - trivial ) homomorphism .
we will be using the equivalent definition @xmath128 of remark r. assume first that @xmath129 is not of the special case excluded from theorem tc , and so @xmath72 is generated by the elements @xmath81 with @xmath77 .
if @xmath130 ( @xmath131 ) then since @xmath132 , induction on lemma p1 implies @xmath133 which clearly implies that @xmath134 is a homomorphism . we are left with the case @xmath135 . by proposition pc2 , @xmath136 where @xmath137 .
we identify @xmath28 with @xmath138 via such an isomorphism .
the set of all elements in @xmath28 with @xmath139 is @xmath140 where @xmath141 and @xmath142 . if @xmath143 and @xmath144 then @xmath145 , whereas if @xmath146 are in the same @xmath147 then @xmath148 .
it follows that any @xmath90 must either map each @xmath147 into itself or map @xmath149 into @xmath150 and @xmath150 into @xmath149 .
so @xmath71 is of the form @xmath151 or @xmath152 where @xmath153 ( @xmath154 . )
such a map will be denoted by @xmath155 or @xmath156 respectively . if @xmath157 then @xmath158 iff @xmath159 and @xmath160 and so @xmath161
so @xmath162 so @xmath163 .
( the @xmath134 on the left is the function on @xmath72 and the @xmath134 on the right is the function on @xmath164 . ) if @xmath165 then @xmath166 iff @xmath167 and @xmath168 that is @xmath169 is of the form @xmath170 with @xmath171 .
and so @xmath172 so @xmath173 . now , since @xmath174 , @xmath175 belongs to the general case , and so we already know @xmath176 .
so we have shown for both @xmath177 and @xmath178 that @xmath179 . now if @xmath180 and @xmath181 then @xmath182 is of the form @xmath183 or @xmath184 . in any case
( again using the fact that @xmath134 on @xmath175 is a homomorphism ) we get that @xmath185 . finally , @xmath124 is not trivial since @xmath186 for any @xmath73 with @xmath77 .
[ r1 ] 1 . for @xmath187 ( the group of @xmath188 orthogonal matrices over @xmath189 )
@xmath190 iff @xmath191 . and
so by remark r , @xmath124 may be thought of as an analogue of the homomorphism @xmath192 .
( @xmath193 on @xmath72 is of course trivial . )
our expression for @xmath134 is meaningful on the whole of @xmath194 , however @xmath134 is in general not a homomorphism on @xmath194 or even on its subgroup @xmath195 of maps preserving @xmath33 . for @xmath196
, we note that though the identification of @xmath28 with @xmath138 in the proof of theorem t1 is not unique , the ( unordered ) pair of sets @xmath197 is uniquely defined by its mentioned properties , namely , @xmath198 , @xmath145 for @xmath199 , and @xmath148 for @xmath200 , @xmath201 .
it follows , as we have noticed , that any @xmath90 either preserves each @xmath147 ( then @xmath157 , ) or interchanges the @xmath147s ( then @xmath165 . )
[ u ] let @xmath202 .
@xmath90 will be called a _ @xmath203-map _ if @xmath71 interchanges @xmath149 and @xmath150 .
[ pum ] let @xmath202 . if @xmath71 is a @xmath203-map such that @xmath204 then @xmath205 .
@xmath206 and so by the proof of theorem t1 , @xmath207 .
let @xmath0 be a closed orientable surface .
@xmath208 from now on will always denote @xmath11 ( considered as a vector space over @xmath27 . )
let @xmath209 be a quadratic form whose associated bilinear form @xmath33 is the algebraic intersection form @xmath210 of @xmath208 .
( in particular , @xmath30 is non - degenerate . )
let @xmath211 denote the mapping class group of @xmath0 i.e. the group of all orientation preserving diffeomorphisms @xmath6 up to isotopy .
for @xmath6 , let @xmath9 denote the map it induces on @xmath208 .
the _ orthogonal mapping class group of @xmath0 with respect to @xmath30 _ will be the subgroup @xmath23 of @xmath211 defined by @xmath212 .
a simple closed curve will be called a _
circle_. if @xmath213 is a circle in @xmath0 , the homology class of @xmath213 in @xmath208 will be denoted by @xmath214 $ ] .
given a circle in @xmath0 , a dehn twist along @xmath213 will be denoted @xmath215 .
( we will not establish a convention as to which of the two possible dehn twists is @xmath215 and which is @xmath216 , rather , it will be clear in each case which of the two should be used . )
the map induced on @xmath208 by @xmath215 is @xmath217}$ ] of definition d0 .
and so by proposition pc4 , @xmath218 iff @xmath219)=1 $ ] or @xmath214=0 $ ] . also , since @xmath220})^2=id$ ] , @xmath221 for any circle @xmath213 . in view of this
we make the following definition : [ d1 ] a map @xmath6 will be called _ good _ if it is of one of the following forms : 1 .
@xmath222 for some circle @xmath213 .
2 . @xmath223 for a circle @xmath213 with @xmath219)=1 $ ] .
@xmath223 for a circle @xmath213 with @xmath214=0 $ ]
. a good map will be called of type 1 , 2 or 3 accordingly .
the purpose of this section is to show that except for the special case when @xmath224 and @xmath80 , @xmath23 is generated by the good maps . for the mentioned special case
, we will show that one more generator is required . whenever we consider two circles in @xmath0
, we will assume that they intersect transversally .
@xmath225 will then denote the number of intersection points between circles @xmath226 and @xmath227 .
( and so the algebraic intersection @xmath228\cdot [ c_2]$ ] in @xmath208 is the reduction mod 2 of @xmath225 . ) given two circles @xmath229 in @xmath0 with @xmath230 , there are two ways for joining them into one circle @xmath213 via surgery at their intersection point .
@xmath213 will be called a _ merge _ of @xmath231 and @xmath232 .
if @xmath231 and @xmath232 are oriented , then @xmath213 will be called the _ positive _ or _ negative _ merge of @xmath231 and @xmath232 , according to whether the surgery is performed so that the orientations of @xmath231 and @xmath232 match or do not match , respectively .
[ l1 ] let @xmath229 be two oriented circles in @xmath0 with @xmath233 , let @xmath234 be their intersection point and let @xmath213 be their negative merge . 1 . @xmath215 followed by an isotopy performed in a thin neighborhood of @xmath235 , maps @xmath231 orientation preservingly onto @xmath232 .
if @xmath236 is another circle passing @xmath234 , and otherwise disjoint form @xmath231 and @xmath232 , and if @xmath231 and @xmath232 cross @xmath236 at @xmath234 in the same direction , then the above dehn twist and isotopy may be performed while fixing @xmath236 .
see fig . f1 .
[ l2 ] let @xmath231 be a circle in @xmath0 and @xmath232 an arc connecting two points of @xmath231 , and whose interior is disjoint from @xmath231 .
assume that at the two endpoints of @xmath232 , @xmath231 passes @xmath232 in the same direction ( as in fig .
let @xmath237 be the two parts of @xmath231 into which it is separated by @xmath238 and let @xmath239 .
then @xmath240 followed by an isotopy performed in a thin neighborhood of @xmath213 , maps @xmath231 onto the circle obtained by surgering @xmath231 along the arc @xmath232 as in fig . f2d .
see fig . f2 .
the setting for the following lemma is that of sections a , b : [ l0 ] assume @xmath129 is not of the two special cases @xmath241 , @xmath80 and @xmath242 , @xmath80 . let @xmath243 , @xmath244 , be independent vectors with @xmath245 and @xmath246 for all @xmath64 .
let @xmath247 ( the subspace spanned by @xmath248 ) and let @xmath249 be two vectors with @xmath250 and @xmath251 .
then there exists @xmath252 with @xmath253 and @xmath254 .
assume first @xmath255 .
we first find @xmath256 such that @xmath257 .
since @xmath258 there are @xmath259 with @xmath260 .
if also @xmath261 or @xmath262 then we have a @xmath232 .
otherwise @xmath263 is our @xmath232 .
if @xmath264 we are done with @xmath265 , otherwise take @xmath266 .
now assume @xmath267 , so @xmath268 and @xmath269 .
if @xmath270 , take @xmath271 with @xmath148 .
if @xmath264 we are done with @xmath265 , otherwise define @xmath272 .
if @xmath273 , take @xmath274 with @xmath275 and @xmath45 ( proposition pc3 ) . if @xmath276 we are done with @xmath277 , otherwise define @xmath278 . in either case @xmath33
is non - degenerate on @xmath203 , and so it is non - degenerate on @xmath279 .
if @xmath280 it follows that @xmath30 can not be identically 0 on @xmath279 .
take any element @xmath281 with @xmath282 then we are done with @xmath283 or @xmath284 respectively .
so we are left with the case @xmath285 . if @xmath241 then we have assumed @xmath286 and so we must have had @xmath264 . if @xmath242 then again we have assumed @xmath286 and so @xmath287 ( since @xmath250 ) so again we must have had @xmath276 .
[ r3 ] when @xmath242 then in the proof of lemma l0 above , we havent used the additional assumption that @xmath286 in the following two cases : ( 1 ) when @xmath255 .
( 2 ) when @xmath267 and @xmath288 ( since then @xmath280 . )
[ l3 ] assume @xmath289 are not of the two special cases @xmath290 and @xmath291 .
let @xmath248 be disjoint circles in @xmath0 with @xmath292)=1 $ ] and such that @xmath293,\dots,[w_k]$ ] are independent in @xmath208 ( which is equivalent to @xmath294 not separating @xmath0 . )
let @xmath295 be oriented circles in @xmath0 with @xmath296)=g([a_2])=1 $ ] and such that @xmath295 are each disjoint from @xmath294 and @xmath297,[a_2]\not\in\langle[w_1],\dots,[w_k]\rangle$ ] .
then there is a sequence @xmath298 of good maps of type 1 and 2 which all fix @xmath294 and such that @xmath299 ( followed by an isotopy fixing @xmath294 ) maps @xmath300 orientation preservingly onto @xmath301 .
assume first that @xmath297\cdot[a_2]=1 $ ] .
if actually @xmath302 then we are done by lemma l1 since the merge @xmath213 of @xmath300 and @xmath301 satisfies @xmath219)=g([a_1])+g([a_2])+[a_1]\cdot[a_2]=1 $ ] and so @xmath215 is a good map of type 2 .
so assume @xmath303 is some odd number @xmath304 .
then necessarily there exist two consecutive crossings along @xmath301 , at which @xmath300 crosses @xmath301 in the same direction .
applying the map @xmath240 of lemma l2 ( @xmath231 is here @xmath300 and @xmath232 is a portion of @xmath301 ) reduces @xmath303 by precisely 2 , and so again @xmath297\cdot[a_2]=1 $ ] and so we may continue by induction .
assume now @xmath297\cdot[a_2]=0 $ ] .
by lemma l0 there is @xmath305 with @xmath306 , @xmath307=x\cdot[a_2]=1 $ ] and @xmath308=0 $ ] for all @xmath2 .
( @xmath309,\dots,[w_k]\rangle$ ] follows . )
there exists a circle @xmath213 in @xmath0 with @xmath214=x$ ] and such that @xmath213 is disjoint from each @xmath310 .
( start with any embedded representative and surger it along the @xmath310s until it is disjoint from all of them .
this is possible since the number of intersection points with each @xmath310 is even .
then connect the various components to each other by surgery .
this is possible since @xmath311 is connected and since there are no orientations to consider . ) by the previous case we may now map @xmath300 onto @xmath213 , and from there , orientation preservingly onto @xmath301 .
[ r4 ] if @xmath289 are of the special case @xmath291 then if in lemma l3 we further assume either that @xmath297\cdot[a_2]=1 $ ] or that @xmath297=[a_2]$ ] or that @xmath255 then it follows from the proof of lemma l3 and from remark r3 , that the conclusion of lemma l3 still holds .
[ um ] let @xmath291 .
a map @xmath312 such that @xmath313 is a @xmath203-map on @xmath208 ( definition u ) will again be called a @xmath203-map .
( such maps clearly exist . )
[ t3 ] if @xmath289 are not of the special case @xmath291 then @xmath23 is generated by the good maps . in the mentioned special case
, @xmath23 is generated by the good maps and any one @xmath203-map .
we first assume we are not in the two special cases appearing in lemma l3 , in particular , we are not in the special case of this theorem .
let @xmath314 and let @xmath43 be circles such that @xmath315 ( @xmath316 ) , @xmath317 and @xmath318)=1 $ ] .
( start with such circles without the assumption on @xmath30 .
then if for some @xmath2 both circles have @xmath319 , replace one of them with their merge .
then exchange @xmath320 with @xmath321 if necessary . )
now let @xmath322 .
we will compose @xmath10 with good maps until ( after isotopy ) we arrive at the identity .
we may first use lemma l3 to bring the @xmath320s one by one back to place .
indeed , after @xmath323 are in place , consider @xmath323 , @xmath324 , @xmath325 , as the @xmath248 , @xmath300 , @xmath301 of lemma l3 , respectively , assigning orientations to @xmath325 and @xmath324 which correspond via @xmath10 .
so assuming @xmath10 fixes all @xmath320s , we bring each @xmath321 back to place , and assume we have already done this for all @xmath326 .
denote @xmath327 , @xmath328 and let @xmath234 be the intersection point of @xmath231 and @xmath232 . since @xmath231 is fixed by @xmath10 , @xmath329 must also pass @xmath234 , and since @xmath10 is orientation preserving , @xmath232 and @xmath329 must cross @xmath231 at @xmath234 in the same direction , ( where orientations on @xmath232 and @xmath329 correspond via @xmath10 . ) our permanent assumption that any circles we consider intersect transversally , may still be maintained for the intersection of @xmath232 and @xmath329 at @xmath234 .
assume first that @xmath234 is the only intersection point between @xmath232 and @xmath329 .
let @xmath213 be the negative merge of @xmath232 and @xmath329 .
@xmath330)=g([b])$ ] ( since @xmath10 preserves @xmath30 ) and so @xmath219)=g([b])+g([h(b)])+[b]\cdot[h(b)]=1 $ ] . by lemma l1 , @xmath215 and
an isotopy bring @xmath329 orientation preservingly onto @xmath232 , and the dehn twist and isotopy may be performed while fixing @xmath231 and while fixing all other @xmath320s and all @xmath321s with @xmath326 .
( our @xmath329 , @xmath232 , @xmath213 and @xmath231 , correspond to @xmath231 , @xmath232 , @xmath213 and @xmath236 of lemma l1 respectively . )
so assume now that there are additional intersection points between @xmath329 and @xmath232 besides @xmath234 .
choose a side of @xmath231 in @xmath0 .
let @xmath331 be the first additional intersection point , when moving from @xmath234 along @xmath232 into the chosen side .
let @xmath332 be a circle parallel and close to @xmath231 on the chosen side .
@xmath333)=1 $ ] .
if @xmath332 is close enough to @xmath231 , then @xmath332 intersects both @xmath232 and @xmath329 each at a single point , and so @xmath334 applied to @xmath329 will add precisely one intersection point between @xmath329 and @xmath232 .
we denote this new intersection point by @xmath335 .
if we choose the orientation of the dehn twist @xmath334 correctly , then @xmath329 will cross @xmath232 at @xmath335 in the same direction that it crosses @xmath232 at @xmath331 .
let @xmath213 be the circle which is the union of the subarcs of @xmath232 and @xmath329 which are defined by @xmath335 and @xmath331 , and which do not contain @xmath234 , and so @xmath213 is disjoint from @xmath231 . by lemma l2
, @xmath240 applied to @xmath329 reduces the intersection between @xmath329 and @xmath232 by precisely 2 .
and so we have first increased the intersection by 1 and then decreased it by 2 , and so we may continue by induction .
we are now in the situation that all @xmath320s and @xmath321s are fixed , so we are left with performing a map fixing all @xmath320s and @xmath321s , which is equivalent to performing a boundary fixing map on a disc with @xmath12 holes .
the group of all such maps is known to be generated by dehn twists .
now , any circle in the complement of the @xmath320s and @xmath321s is bounding in @xmath0 , and so these dehn twists are good maps of type 3 .
this completes the proof for the general case .
the case @xmath290 does not rely on the above , and will not be used in the sequel .
we differ it to the end of section d1 .
we are left with the case @xmath291 .
we will show that by adding any @xmath203-map @xmath336 , the above proof can be made to work . by remark r4 ,
the only problem we have is when moving the first circle @xmath300 .
let @xmath197 be the pair of subsets of @xmath208 from the definition of @xmath203-map , and say @xmath297\in v_1 $ ] .
if @xmath337\in v_2 $ ] then @xmath338\in v_1 $ ] , and so we may assume @xmath297 $ ] and @xmath337 $ ] are both in @xmath149 .
but then either @xmath297=[h(a_1)]$ ] or @xmath297\cdot [ h(a_1)]=1 $ ] . by remark
r4 again , lemma l3 applies , and so the above process works .
[ r5 ] if @xmath339 denotes the map @xmath340 then @xmath341 is known to be @xmath342 , the group of maps preserving the intersection form on @xmath208 . since @xmath343 and @xmath344 , @xmath345 is onto .
since @xmath33 is unique up to isomorphism for every given dimension @xmath54 , we see that our theorem t3 implies theorem tc ( the cartan - dieudonne theorem for the field @xmath27 . )
let @xmath346 be an annulus .
there are two regular homotopy classes of immersions @xmath347 .
this follows from the smale - hirsch theorem ( [ h ] ) and the fact that @xmath348 . the first class is represented by an embedding whose image lies in a plane in @xmath349 , and the second class , by an embedding that differs from the former by one full twist .
for an immersion @xmath347 we let @xmath350 be 0 or 1 according to whether @xmath2 belongs to the first or second class , respectively .
let @xmath0 be a closed orientable surface , and @xmath5 an immersion , then @xmath2 determines a quadratic form @xmath354 whose associated bilinear form is the intersection form on @xmath208 , as follows : let @xmath305 , choose a circle @xmath213 in @xmath0 such that @xmath214=x$ ] and define @xmath355 .
one needs to verify that @xmath356 is independent of the choice of @xmath213 and that @xmath357 .
this has been done in [ p ] in the more general setting of surfaces which are not necessarily orientable .
( it is then necessary for the quadratic form to attain values in @xmath358 rather than @xmath27 , to accommodate for the half twists of mobious bands , and the arf invariant attains values in @xmath359 . ) for immersions @xmath1 , @xmath360 will denote that @xmath2 and @xmath3 are regularly homotopic in @xmath349 .
the following has been shown in [ p ] : let @xmath18 be the group of _ all _ diffeomorphisms of @xmath0 ( not necessarily orientation preserving ) up to isotopy . given a quadratic form @xmath30 on @xmath208 ( whose associated bilinear form is the intersection form )
let @xmath22 be the subgroup of @xmath18 defined by @xmath365 .
( @xmath23 is then a subgroup of index 2 in @xmath22 . )
now let @xmath5 be an immersion and @xmath6 a diffeomorphism . by theorem pi1(1 ) , @xmath366 iff @xmath367 .
it is easy to see that @xmath368 and so we get : [ qii ] let @xmath1 be two regularly homotopic generic immersions .
we define @xmath4 by @xmath375 , where @xmath372 is any generic regular homotopy between @xmath2 and @xmath3 .
this is well defined by theorem q. if @xmath376 are two regular homotopies such that the final immersion of @xmath372 is the initial immersion of @xmath377 , then @xmath378 will denote the regular homotopy that performs @xmath372 and then @xmath377 .
recall that for @xmath96 we have defined @xmath387 , and have shown that @xmath124 is a homomorphism ( theorem t1 ) .
for @xmath388 let @xmath389 be 0 or 1 according to whether @xmath10 is orientation preserving or reversing , respectively , and let @xmath314 .
since @xmath390 and @xmath340 are also homomorphisms , the following @xmath391 is a homomorphism : @xmath392 let @xmath5 be an immersion and let @xmath213 be a circle in @xmath0 such that @xmath213 is disjoint from the multiplicity set of @xmath2 . adding a _ ring _ to @xmath2 along @xmath213 will mean to change @xmath2 into a new immersion @xmath3 in the following way : let @xmath395 , thin enough so that @xmath396 is an annulus which is still disjoint from the multiplicity set .
let @xmath397 and @xmath231 be a disc and an arc .
let @xmath398 be a proper embedding and let @xmath399 be a proper immersion with one transverse intersection and such that @xmath400 .
parametrize @xmath203 and @xmath346 as @xmath401 and @xmath402 so that @xmath403 will be given by @xmath404 .
now , the new immersion @xmath3 will be given by @xmath405 on @xmath346 , and will be the same as @xmath2 outside @xmath346 .
there are basically two ways for adding a ring to @xmath2 along @xmath213 , depending on what side of @xmath346 in @xmath349 the ring will be facing ( which in turn depends on our choice of @xmath399 . )
if @xmath2 is an embedding , then @xmath406 separates @xmath349 into two pieces , one compact and one non - compact .
they will be denoted @xmath407 and @xmath408 respectively .
and so if @xmath2 is an embedding then we have a natural way for distinguishing the two possibilities for adding a ring along a given circle @xmath213 , namely , the ring is facing either @xmath407 or @xmath408 .
note that @xmath409 are homotopic relative @xmath410 ( but not regularly homotopic . ) and so if @xmath411 with @xmath412 , then @xmath2 and @xmath3 are homotopic in @xmath413 ( but in general not regularly homotopic . )
we now present two moves on immersions , that have been introduced in [ n ] .
let @xmath414 be an annulus and @xmath415 be a disc . move @xmath346 ( resp .
@xmath39 ) is a regular homotopy which is applied to a proper immersion of @xmath414 ( resp .
@xmath415 ) into a ball @xmath416 and which fixes @xmath417 .
move @xmath346 begins with the standard embedding of @xmath414 in @xmath416 , and adds a ring and a dehn twist along parallel essential circles in @xmath414 .
the ring may face either side of @xmath414 in @xmath416 and the dehn twist may have either orientation .
f3 shows one of the possibilities .
the reverse move , going from right to left in fig .
f3 , will be called an @xmath418 move .
move @xmath39 is described in fig .
it begins with a specific immersion of @xmath415 , with two arcs of intersection , and replaces them with two other arcs of intersection .
it is easy to see ( and has been shown in [ n ] ) that the initial and final immersions that we have presented for the @xmath346 and @xmath39 moves , are indeed regularly homotopic in @xmath416 ( while keeping @xmath419 fixed . )
move @xmath346 ( resp .
@xmath39 ) will be applied to an immersion @xmath5 inside a ball @xmath416 in @xmath349 for which @xmath420 is an annulus ( resp .
disc ) and @xmath421 is as above .
( the rest of @xmath0 will be kept fixed . ) in particular , an @xmath346 move may be applied to a neighborhood of a circle @xmath213 in @xmath0 iff @xmath213 is disjoint from the multiplicity set of @xmath2 and there is an embedded disc @xmath397 in @xmath349 such that @xmath422 .
the move will then be performed in a thin @xmath423 .
if the circle along which we perform an @xmath346 move happens to bound a disc in @xmath0 , then the dehn twist that is produced is trivial , and may be cancelled by rotating this disc .
the following has been shown in [ n ] : [ rbb ] after we prove theorem main1 , we will know that the assumption @xmath360 in lemma bb is actually unnecessary , as long as @xmath439 .
this is so since @xmath440 does not depend on @xmath2 .
( it is a function of @xmath10 only . )
we begin the proof of theorem main1 .
first let @xmath441 .
by lemma bb ( and since all immersions of @xmath442 in @xmath349 are regularly homotopic ) we may assume @xmath2 is an embedding onto the unit sphere . let @xmath443 be the reflection with respect to the @xmath444-plane , and @xmath19 be such that @xmath445 , then @xmath10 is the unique non - trivial element of @xmath446 and @xmath447 .
the following is a regular homotopy from @xmath2 to @xmath7 : perform an @xmath346 move on the equator of the sphere , such that the ring formed will be facing @xmath408 .
then exchange the northern and southern hemispheres , arriving at @xmath7 .
by proposition pn1 the @xmath346 move contributed 1 mod 2 quadruple points , then exchanging the hemispheres involved only double curves , and so @xmath448 . we proceed by induction on @xmath449 beginning with @xmath450 . for the starting point @xmath450 , we will need to separate the cases @xmath451 and @xmath452 .
this has basically been done in [ n ] .
we present it here with slight variation .
let @xmath71 denote the torus .
we will say an embedding @xmath453 is _ standard _ if its image is the torus @xmath454 obtained by rotating the circle @xmath455 around the @xmath456-axis .
let @xmath457 be the circles in @xmath458 given by @xmath459 and @xmath460 and choose some orientations for @xmath461 and @xmath462 .
for a standard embedding @xmath453 , let @xmath463 and @xmath464 denote the oriented circles in @xmath71 such that @xmath465 and @xmath466 . since @xmath451 then by theorem pi1(2,3 ) @xmath2 is regularly homotopic to a standard embedding .
( take an arbitrary standard embedding @xmath3 , then @xmath362 for some diffeomorphism @xmath467 , but @xmath437 is again a standard embedding . )
so by lemma bb we may assume @xmath2 itself is a standard embedding . by viewing @xmath468 as the basis for @xmath469 ( note @xmath470 coefficients )
we identify @xmath18 with @xmath471 .
we will think of any @xmath19 both as a map from @xmath0 to @xmath0 and as a @xmath472 matrix . if @xmath19 then @xmath473 ( now @xmath27 coefficients ) presented with respect to the basis @xmath474,[l]$ ]
is simply the @xmath27 reduction of the matrix @xmath10 .
@xmath475)=g^i([l])=0 $ ] and for @xmath476+[l]$ ] , @xmath477 .
a matrix @xmath478 is in @xmath20 if @xmath9 preserves @xmath21 .
this will happen iff @xmath479 and @xmath480,[l]\})=\{[m],[l]\}$ ] , i.e. iff the @xmath27 reduction of @xmath10 is either @xmath481 $ ] or @xmath482 $ ] . by means of row and column operations one can show that this subgroup of @xmath471 is generated by the following four elements : @xmath483 , @xmath484 , @xmath485 , @xmath486 .
since the @xmath27 reduction of @xmath487 is @xmath488 and that of @xmath489 is @xmath490 , and since @xmath491 , @xmath492 and @xmath493 , we have @xmath494 and @xmath495 .
we will now show that the values of @xmath496 , @xmath497 are the same .
@xmath498 and @xmath499 are @xmath500 and @xmath501 respectively .
let @xmath397 be a compressing disc for @xmath502 in @xmath349 and @xmath503 thin enough so that @xmath504 is a standard annulus in @xmath39 .
in @xmath39 we isotope @xmath504 to be a thin tube , we then perform the `` belt trick '' on this tube ( fig .
f5 ) and then isotope @xmath504 back to place .
the effect of this regular homotopy is precisely @xmath500 and it involves only double curves , and so @xmath505 . in the same way @xmath506 . up to isotopy in @xmath71 , @xmath507 where @xmath508 is the reflection of @xmath349 with respect to the @xmath444-plane
this may be achieved by a regular homotopy similar to the one we have used for the case @xmath441 , as follows : perform an @xmath346 move on each of the two circles @xmath509 and @xmath510 , so that the ring formed by each of them is facing @xmath408 and such that the two dehn twists formed will have opposite orientations and so will cancel each other .
( the @xmath346 moves are performed inside thin neighborhoods of compressing discs for the two circles . )
so we remain with just the two rings .
we may now exchange the upper and lower halves of @xmath71 until we arrive at @xmath511 .
the two @xmath346 moves each contributed @xmath512 quadruple points and the final stage involved only double curves , and so all together @xmath513 . for @xmath489 , isotope @xmath71 until it has the shape of a large sphere with a tiny handle at its north pole .
now exchange the northern and southern hemispheres .
this will involve only double curves , and will result in a sphere having a tiny handle at the south pole and a ring along the equator .
we cancel this ring with an @xmath418 move in a thin neighborhood of the plane of the equator , resulting in an embedding again .
we may think of @xmath514 and @xmath515 as being contained in the tiny handle , and so tiny compressing discs for @xmath514 and @xmath515 may be pulled along with the regular homotopy .
the compressing disc of @xmath514 now lies in @xmath408 , and the compressing disc of @xmath515 now lies in @xmath407 .
it follows that the final embedding may be isotoped to a standard embedding i.e. to a map of the form @xmath7 , and this @xmath10 is orientation reversing and exchanges @xmath514 and @xmath515 .
and so @xmath10 must be either @xmath516 or @xmath517 .
these two maps composed with @xmath2 are isotopic in @xmath349 ( via a half revolution about the @xmath518-axis ) and so we may assume @xmath519 .
our regular homotopy had a portion involving just double curves , then one @xmath418 move , and finally some isotopy , and so @xmath520 .
this completes the proof that @xmath521 for every @xmath393 for @xmath0 a torus and @xmath451 .
we now give the promised completion of the proof of theorem t3 .
we need to show that @xmath23 is generated by good maps when @xmath290 .
choose two oriented circles @xmath229 with @xmath522)=g([b])=0 $ ] and @xmath230 as a basis for @xmath469 , thus identifying @xmath211 with @xmath523 .
again we see @xmath524 is in @xmath23 iff its @xmath27 reduction is either @xmath488 or @xmath490 . by row and column operations
, we then see that @xmath23 is generated by @xmath483 , @xmath484 and @xmath525 .
now @xmath498 and @xmath499 are @xmath526 and @xmath527
. if @xmath213 is the positive merge of @xmath231 and @xmath232 then @xmath219)=1 $ ] and @xmath215 is @xmath528 .
since @xmath529 we see @xmath23 is generated by @xmath526 , @xmath527 and @xmath215 . by theorem pi1(1 ) , @xmath2 is regularly homotopic to an immersion which is obtained from a standard embedding @xmath453 by adding a ring along the circle @xmath213 which is the positive merge of @xmath463 and @xmath464 , and such that the ring is facing @xmath530 .
( one checks directly that such an immersion @xmath3 has @xmath531 , but then @xmath17 since on @xmath28 of dimension 2 there is only one @xmath30 with @xmath286 . ) by lemma bb we may assume @xmath2 itself is this new immersion ( fig .
since all non - zero elements in @xmath208 have @xmath532 , it follows that @xmath533 and so @xmath534 .
since @xmath27 is abelian , it is enough to verify @xmath521 only on normal generators , and we claim that @xmath535 and @xmath536 are normal generators of @xmath18 where the identification with @xmath471 is via @xmath537 .
( there are no @xmath538 since such are defined only for standard embeddings . ) indeed , @xmath539 which as we have noticed in the end of section d1 , is a dehn twist along @xmath213 ( the positive merge of @xmath463 and @xmath464 . ) now , any dehn twist is a normal generator of @xmath211 , and since @xmath540 are orientation reversing , they normally generate the whole of @xmath18 .
as above , we see that @xmath541 and @xmath542 .
the regular homotopy we construct for @xmath543 is as follows : let the ring become thicker , and at the same time let the `` body '' of the torus become thinner , until they are equal in width , ( fig .
f6a@xmath544b . )
now using the intersection circle as an axis of rotation , we perform a half revolution of the torus around it , interchanging the two equal - width rings , ( fig . f6b@xmath544c . )
we then return to our original position by reversing our first step .
( fig . f6c@xmath544d . )
we thus arrive at an immersion of the form @xmath7 .
the map @xmath467 may best be understood by looking at the intermediate stage of the regular homotopy , fig .
we see here that @xmath10 maps @xmath463 to @xmath545 and @xmath213 to itself .
i.e. the column @xmath546 to @xmath547 and @xmath548 to itself .
and so indeed @xmath549 .
we had no quadruple points ( actually no singular occurrences at all ) and so @xmath550 . for @xmath551 , we first imitate the regular homotopy we have had for @xmath552 of section d1 i.e. we perform that regular homotopy on @xmath553 and carry the ring along . if before exchanging the upper and lower hemispheres we make sure that the ring is situated at the tiny handle , then this exchange will have at most triple points , and the ring will not interfere with the @xmath418 move , and so at the end of this process we will have @xmath554 .
the immersion @xmath555 we arrive at , is the immersion obtained by adding a ring @xmath556 to the embedding @xmath557 along the circle @xmath558 , that is the same circle along which @xmath556 was originally situated , but now @xmath556 is facing @xmath559 instead of @xmath530 and so @xmath555 is not of the form @xmath7 .
we fix this with a regular homotopy as follows : perform an @xmath346 move on a little disc bounding circle in @xmath71 near @xmath556 , forming a ring @xmath560 facing @xmath530 . see fig . f7a .
( the dotted line in fig .
f7 is the intersection curve of @xmath556 .
@xmath556 itself is not seen since it is facing @xmath559 . )
we then elongate @xmath560 along side @xmath556 , until it approaches itself from all the way around .
f7a@xmath544b@xmath544c . )
we then perform a @xmath39 move , turning @xmath560 into two rings which are parallel to @xmath556 , but facing @xmath530 .
( fig . f7c@xmath544d . )
it is then easy to construct an explicit regular homotopy so that @xmath556 and the new ring which is adjacent to it , will cancel each other , and with no quadruple points at all .
( the idea is as follows : let @xmath561 be a proper immersion of an arc @xmath231 into a disc @xmath397 with two loops facing opposite sides , as in the front disc of fig .
f8a . there is a regular homotopy @xmath562 fixing @xmath563 , from @xmath564 to an embedding as in the front disc of fig .
if @xmath565 is generic then it has at most triple points .
now the regular homotopy @xmath566 is a regular homotopy which begins with a pair of rings facing opposite sides , and cancels them .
f8 depicts a portion of @xmath402 in the initial and final immersions . indeed since @xmath565 has at most triple points , so will @xmath567 , but @xmath567 is not generic .
nevertheless , any specific @xmath567 may serve as a guide for constructing an explicit generic regular homotopy with no quadruple points , and which begins and ends with the same immersions . ) since @xmath556 and one of the new rings have disappeared , we are left with one ring facing @xmath530 and situated along a circle parallel to @xmath568 . by pushing it precisely to @xmath568 we finally get an immersion of the form @xmath7 .
the regular homotopy which started with @xmath555 and replaced the ring @xmath556 with a ring facing @xmath530 , took place inside some @xmath569 , and so the homotopy class into @xmath413 is still that of @xmath557 and so ( up to isotopy in @xmath0 ) the new immersion is indeed @xmath570 .
( @xmath571 is a homotopy equivalence , and so two diffeomorphisms @xmath572 are isotopic in @xmath0 iff @xmath573 are homotopic in @xmath413 . ) finally , our regular homotopy from @xmath2 to @xmath555 involved @xmath512 quadruple points , then from @xmath555 to @xmath570 we had one @xmath346 move , one @xmath39 move , and a regular homotopy with no quadruple points , and so all together indeed @xmath574 .
assume @xmath575 .
by theorem t3 @xmath26 is generated by dehn twists and squares of dehn twists , and in the special case @xmath576 we also need a @xmath203-map .
other than the special @xmath203-map generator , which will be dealt with last , each generator @xmath10 fixes all but a regular neighborhood of a circle @xmath231 ( and we may assume @xmath231 does not bound a disc in @xmath0 . ) if @xmath231 is non - separating then there is a circle @xmath232 in @xmath0 with @xmath230 .
@xmath577 is a punctured torus , and so @xmath578 is a circle separating @xmath0 into two subsurfaces @xmath579 of smaller genus than @xmath0 and with @xmath580 . if @xmath231 is separating , then a nearby parallel circle @xmath213 will again separate @xmath0 in this way .
now @xmath20 needs one additional generator .
choosing any separating circle @xmath213 in @xmath0 there is clearly an orientation reversing @xmath6 which preserves the two sides of @xmath213 and which induces the identity on @xmath208 and so @xmath393 . and so finally we have a set of generators for @xmath20 , each of which ( except for the @xmath203-map ) preserves such a separation of @xmath0 by a circle @xmath213 into @xmath579 of smaller genus .
let @xmath581 . slightly diminishing @xmath579 to be the components of @xmath582 , we may still assume @xmath583 , @xmath201 .
since @xmath213 is separating in @xmath0 , @xmath584)=0 $ ] and so @xmath585 is regularly homotopic to a standard embedding of @xmath346 , in the shape of a thin tube . by means of [ s ] ( namely the proof of theorem 2.1 )
we may extend such a regular homotopy of @xmath346 to the whole of @xmath0 .
we now stretch this tube to be very long , at the same time pulling @xmath586 and @xmath587 rigidly away from each other until they are disjoint .
see fig . f9a . by taking a smaller @xmath346 if necessary
, we may assume @xmath588 is disjoint from @xmath589 , fig . f9b . by lemma bb
, we may assume that this is in fact our immersion @xmath2 .
let @xmath590 be the closed surfaces obtained by gluing a disc @xmath591 to @xmath592 and let @xmath593 be an extension of @xmath594 .
if the tube @xmath588 is very thin , then there is also a naturally defined extension @xmath595 of @xmath596 .
we may further assume that the thin ball @xmath39 in @xmath349 which is bounded by the sphere @xmath597 , is disjoint from @xmath589 .
since @xmath598 preserves @xmath599 then @xmath600 preserves @xmath601 .
it follows that there is a regular homotopy @xmath382 between @xmath602 and @xmath603 .
we perform @xmath604 and @xmath605 inside disjoint balls , and we let the thin tube @xmath346 be carried along . if we make sure no quadruple points occur in @xmath606 and @xmath607 , and the thin tube @xmath346 does not pass triple points , then the regular homotopy @xmath372 induced on @xmath0 in this way will have the sum of the number of quadruple points of @xmath604 and @xmath605 .
now if @xmath10 is orientation preserving then so are @xmath600 , in particular @xmath608 is orientation preserving .
so if we had carried the thin ball @xmath39 along with the tube @xmath346 , then it would now approach the @xmath591s from the same side it had for @xmath2 .
and so we may continue the regular homotopy on the tube @xmath346 , still not passing through triple points , and cancelling all knotting by having the thin tube pass itself , until it is back to its original place , and this will not contribute any quadruple points .
however , the new embedding of @xmath346 may differ from @xmath609 by some number of dehn twists as in fig .
we may resolve this by rigidly rotating say @xmath586 around the axis of the tube .
if on the other hand @xmath10 is orientation reversing , then after applying @xmath604 and @xmath605 and carrying the tube along , the thin ball @xmath39 will approach both @xmath591s from the wrong side . and
so after we cancel all knotting , the tube @xmath346 will be as in fig f9d .
f10 presents a regular homotopy that resolves this , and has @xmath512 quadruple points .
f10a depicts the relevant part of fig .
f9d , where the regular homotopy will take place . fig .
f10a@xmath544b@xmath544c is a regular homotopy with no singular occurrences , or alternatively may be thought of as an ambient isotopy of @xmath349 .
it shows that we may view the immersion of @xmath346 as a sphere with two rings facing outward , each of which has a tube attached to it .
we now perform a @xmath39 move which joins the two rings into one ring with two tubes attached to it , fig .
f10c@xmath544d . again by ambient isotopy , the ring may be brought to the equator , fig .
finally we exchange the northern and southern halves of the sphere , arriving at an embedding , fig .
this regular homotopy involved a @xmath39 move and a portion involving only double curves , and so indeed it had @xmath512 quadruple points .
we then continue to bring @xmath346 back to place . as
above , the new embedding of @xmath346 may differ from @xmath610 by some number of dehn twists , and those may be cancelled by rigidly rotating @xmath586 . we have thus constructed a regular homotopy between @xmath2 and @xmath7 such that the number mod 2 of quadruple points , is the sum of the number occurring in the @xmath611s in case @xmath10 is orientation preserving , and the sum plus 1 , in case @xmath10 is orientation reversing .
in other words @xmath612 . by the induction hypothesis , @xmath613 , @xmath201 .
let @xmath614 and notice @xmath615 , @xmath616 and @xmath617 and so @xmath618 . so finally : @xmath619 .
we now deal with the special @xmath203-map generator which appears when @xmath576 .
since @xmath451 , then by theorem pi1(2,3 ) and lemma bb as before , we may assume @xmath2 is an embedding whose image is two embedded tori connected to each other with a tube , and such that a half revolution around some line in @xmath349 maps @xmath406 onto itself and interchanges the two tori .
let @xmath6 be the map such that @xmath7 is the final embedding of this half revolution , then @xmath10 is orientation preserving and so @xmath620 .
take a circle @xmath213 in one of the tori with @xmath584)=1 $ ] , then @xmath621 lies in the other torus , and so @xmath214\neq h_*([c])$ ] and @xmath214\cdot h_*([c])=0 $ ] .
it follows that @xmath214 $ ] and @xmath622)$ ] are not in the same @xmath147 of the definition of @xmath203-map , so @xmath9 must be a @xmath203-map , and so @xmath10 is indeed a @xmath203-map on @xmath0 . since @xmath623 then by lemma pum , @xmath624 and so @xmath625 . since there is a rigid rotation between @xmath2 and @xmath7 , @xmath626 .
this completes the proof of theorem main1 .
chevalley : `` the algebraic theory of spinors . ''
columbia university press 1954 . also reprinted in : c. chevalley : `` the algebraic theory of spinors and clifford algebras , collected works , vol .
2 '' springer - verlag 1997 . | let @xmath0 be a closed orientable surface .
if @xmath1 are two regularly homotopic generic immersions , then it has been shown in [ n ] that all generic regular homotopies between @xmath2 and @xmath3 have the same number mod 2 of quadruple points .
we denote this number by @xmath4 .
we show that for any generic immersion @xmath5 and any diffeomorphism @xmath6 such that @xmath2 and @xmath7 are regularly homotopic , @xmath8 where @xmath9 is the map induced by @xmath10 on @xmath11 , @xmath12 is the genus of @xmath0 and @xmath13 is 0 or 1 according to whether @xmath10 is orientation preserving or reversing , respectively
. [ section ] [ thm]lemma [ thm]proposition [ thm]corollary [ thm]definition [ thm]remark |
the combination of an electronic interaction and a strong magnetic field in a two - dimensional electron system yields a rich variety of phases .
these are best classified by the filling factor @xmath6 , which is the electron density divided by the density of a completely filled landau level . in this work
we mainly consider energy calculations on systems with filling factors @xmath7 and @xmath3 .
these system are most suitably described by the chern - simons theory . in solid state physics
this theory is mainly used in the fields of the fractional quantum hall effect and high temperature superconductivity .
the applications of the chern - simons theory in the field of high - temperature superconductivity is based on a work of polyakov @xcite .
since the discovery of the fractional quantum hall effect by tsui , strmer and gossard ( 1982 ) @xcite there were many attempts to explain this experimental observation .
the current chern - simons type theories of this effect are mainly based on a work of jain ( 1989 ) @xcite . in his theory
he mapped the wave functions of the integer quantum hall effect to wave functions of the fractional quantum hall effect . in the case of filling fraction @xmath0
every electron gets two magnetic flux quantums through this mapping . by this transformation
one gets new quasi - particles ( composite fermions ) which do not see any magnetic field in first approximation ( mean field ) .
a field theoretical language for this scenario was first established by halperin , lee , read ( hlr ) ( 1992 ) @xcite as well as kalmeyer and zhang ( 1992 ) @xcite for the @xmath0 system .
the interpretation of many experiments supports this composite fermion picture .
we mention transport experiments with quantum ( anti- ) dots @xcite , and focusing experiments @xcite here .
an overview of further experiments can be found in @xcite .
+ hlr studied many physical quantities within the random - phase approximation ( rpa ) .
most prominent among these is the effective mass of the composite fermions which they found to diverge at the fermi surface @xcite .
recently , shankar and murthy @xcite proposed a new theory of the @xmath0 system .
based upon a transformation of the chern - simons hamiltonian one achieves a separation of the magneto - plasmon oscillators from the total interaction of the system . after restricting the number of the magneto - plasmon oscillators to the number of electrons they got a finite quasi - particle mass which scales with the inverse of the strength of the coulomb repulsion . in this derivation
they calculate a smaller number of self energy feynman graphs than in the rpa .
just recently stern et al .
@xcite calculated the self energy feynman graphs for the theory of shankar and murthy in rpa and got the same divergence of the effective mass as hlr . besides the theory of shankar and murthy there are other alternate formulations of the composite fermionic picture which are formulated entirely in the lowest landau level @xcite .
+ in the following we will use mainly the chern - simons theory of hlr to calculate ground state energies of the @xmath0 and the @xmath3 system . to get composite fermions one has to transform the electronic wave function with a rather singular transformation . due to this transformation one
gets a density - density interaction of the form @xmath8 for small wave vectors . because of this singular interaction the singular diagrams in leading order should be resummed .
these diagrams are well known as the rpa in the jellium problem in @xmath9 . we have shown in an earlier publication @xcite that despite of resumming these singular diagrams one gets an ir - divergence in the canonical potential of the @xmath0 system in rpa due to the @xmath8 interaction .
thereby we used the chern - simons path integral of hlr for the resummation of the diagrams .
we obtained further that the coulomb part of the energy is finite and in good agreement with the coulomb energy of numerical simulations of interacting electrons in the lowest landau level by morf and dambrumenil @xcite as well as girlich @xcite .
we have to mention that c. conti and t. chakraborty @xcite got an excellent agreement with the numerical results by calculating the coulomb energy with the help of the stls theory @xcite which is a generalisation of the rpa theory . in the following it
is shown that we obtained an ir divergence of the magnetic energy in our earlier work because the normal ordering of the chern - simons hamiltonian was not properly taken into account in the path integral of hlr . in @xcite
we correct this error with the help of an intermediate time technique in the path integral and a change of the coupling between the fermion fields and the bosonic fields .
one can show that this path integral has a finite energy in the rpa .
details of the analysis of the rpa for this path integral will be published elsewhere . in the case of the coulomb gas ,
the hartree - fock graphs belong to the maximal divergent feynman graphs @xcite .
this is no longer valid for the chern - simons theory .
for this theory we will get a finite energy for the spin polarized @xmath0 system by calculating the energy of all feynman - graphs which belong either to the hartree - fock graphs or to the maximal divergent graphs ( for a given number of interaction vertices ) . with the help of this principle
one gets a well - suited finite approximation of the magnetic energy .
furthermore , we get the same coulomb energy as in @xcite . + in the second part of this work we will calculate the coulomb energy of the @xmath0 and the @xmath3 system including the spin degree of freedom .
the @xmath3 system is of theoretical interest because it consists of one landau level filled with spin up and spin down electrons . like in the @xmath0 system
the second landau level is half filled with electrons .
so the @xmath10 system should have similar physical properties as the @xmath0 system . as a matter of fact
this is not the case .
eisenstein et al.@xcite were able to show with the help of tilted field experiments that the ground state of the @xmath3 system is spin unpolarized and incompressible .
this is different to the ground state of the @xmath0 system , which is spin polarized and compressible .
numerical simulations of the ground state of the @xmath0 and the @xmath3 system @xcite show that one gets a transition from a spin unpolarized ground state to a spin polarized ground state depending on the strength of the interaction between the electrons . for calculating this transition with perturbational methods
we will construct a chern - simons theory of the @xmath4 systems where @xmath11 is an even number by transforming the coulomb interaction of the second landau level to the first landau level . by neglecting the lowest landau level
, the @xmath3 system behaves like a @xmath0 system with an altered interaction potential . because of this we can calculate the coulomb part of the energy of the @xmath3 system within the rpa formalism of the @xmath0 system .
it will be shown that one gets a transition from a spin polarized to a spin unpolarized ground state for the @xmath0 system as well as for the @xmath12 system depending on the interaction strength .
this transition is in qualitative agreement with the numerical simulations @xcite .
furthermore , we will use the chern - simons theory of the @xmath4 systems to calculate the compressibility , the effective mass and the excitations . + as mentioned above we will treat the energy problem of the @xmath0 and the @xmath3 chern - simons system in rpa .
the use of this approximation is motivated by the similarity of the interaction potentials of these two systems to the one of the @xmath13 jellium problem .
it is well known that for the latter a perturbational calculation of the ground state energy results in a good approximation especially for small densities ( which is a small parameter in this theory ) .
such a small parameter is not existent in the @xmath0 chern - simons theory .
until now it is not clear wether the rpa calculations of the ground state energy in chern - simons theories are in agreement with the experiments .
one aim of this paper is to make one step further to a positive answer to this question .
+ the paper is organized as follows : in section ii we reconsider the polarized @xmath0 system with the help of a path integral which takes the normal ordering of the chern - simons hamiltonian into account .
we calculate the energy of the maximal divergent graphs together with the hartree - fock feynman graphs of the chern - simons hamiltonian . in section iii
we formulate the rpa theory of the @xmath14 chern - simons theory subjected to a spin constraint . in section
iv we formulate a chern - simons theory of the @xmath4 systems and calculate the compressibility , the effective mass and the excitations in rpa . in section
v we calculate the coulomb ground state energy of the spin polarized as well as the spin unpolarized @xmath0 and @xmath3 systems .
in this section we consider interacting spin polarized electrons moving in two dimension in a strong magnetic field @xmath15 directed in the positive @xmath16-direction of the system .
the electronic density of the system is chosen such that the lowest landau level of a non - interacting system is filled to a fraction @xmath17 where @xmath11 is an even number .
we are mainly interested in @xmath18 . after performing a chern - simons transformation @xcite of the electronic wave function one
gets the hamiltonian of the composite fermions as : @xmath19 .
\nonumber \end{aligned}\ ] ] the chern - simons operator @xmath20 is defined by @xmath21 . here @xmath22 creates ( and @xmath23 annihilates ) a composite fermion with coordinate @xmath24 .
@xmath25 is the coulomb interaction where @xmath26 .
@xmath27 is the charge of the electrons and @xmath28 is the dielectric constant of the background field @xmath29 .
@xmath30 is the vector potential @xmath31 and @xmath32 is a homogeneous magnetic field in z - direction @xmath33 where @xmath34 is the unit vector in @xmath35-direction .
we suppose throughout this paper that @xmath15 is a positive number . the function @xmath36 is given by @xmath37 .
we used the convention @xmath38 and @xmath39 in the above formula ( [ 1050 ] ) .
furthermore , we set @xmath40 for the coupling of the magnetic potential to the electrons .
it is well known , that the partition function of the hamiltonian ( [ 1050 ] ) can be written in an operator formalism @xcite with the help of the bosonic chern - simons fields ( @xmath41 ) .
this is shown in the first quantized path integral language .
using standard methods @xcite we can transform this partition function to a path integral .
this path integral is written as @xmath42\ , { \cal d}[\psi^*_l,\psi_l ] { \label{1150 } } \\ & & \times \exp\left[-\epsilon\left ( l_{l}+l_{cs , l}+l_{ee , l}+l^0_{ee , l}\right)\right ] \nonumber \;. \end{aligned}\ ] ] the various functions in ( [ 1150 ] ) are given by @xmath43 together with the norm @xmath44 \exp\left[-\epsilon\left(l_{cs , l}+l^0_{ee , l}\right)\right ] \;.{\label{1175}}\ ] ] the path integral ( [ 1150 ] ) is correct under the gauge condition @xmath45 .
@xmath46 is a imaginary field . @xmath47 and @xmath48 are real fields .
the time slices @xmath28 are defined as @xmath49 .
the index @xmath50 counts the discrete time slices .
furthermore , we have anti - periodic boundary conditions @xmath51 for the grassmann fields @xcite .
the action of the path integral ( [ 1150 ] ) is given by a fermionic term @xmath52 , a bosonic term @xmath53 of the chern - simons form , and two coulomb interaction terms @xmath54 , @xmath55 . in comparison to the chern - simons path integral of hlr @xcite which was used in our earlier publication @xcite we get an additional term
proportional to @xmath56 in @xmath57 ( [ 1160 ] ) .
this term is mainly due to the normal - ordering of the @xmath58 term in the chern - simons hamiltonian @xmath59 ( [ 1050 ] ) .
this is most easily seen by integrating the path integral ( [ 1150 ] ) over the chern - simons fields . due to the additive term
one can not perform the formal limit @xmath60 in ( [ 1150 ] ) .
this limit has to be taken in every additive term after integration over the chern - simons and fermionic fields .
furthermore , we have to remark , that we treat the same formalism as above in order to derive the path integral of shankar and murthy @xcite . there
we get no additional term in comparison to their path integral @xcite . + in the following we integrate ( [ 1150 ] ) over the fermionic fields . using a mean field expansion of the result up to the second order in the bosonic fields one gets after integration for the grand - canonical potential @xmath61 .
@xmath62 is the grand - canonical potential of the coulomb - free electron gas .
we now split @xmath63 into a variety of terms for which one can take the limit @xmath64 rather easily @xcite . @xmath65 the expression @xmath66 corresponds to the rpa graphs which do not belong to the hartree - fock graphs .
@xmath67 are the coulomb and chern - simons fock diagrams .
@xmath68 are the hartree diagrams .
we will see below that the hartree term is due to the normal ordering of the @xmath58-term in ( [ 1050 ] ) . for temperature @xmath69 , @xmath66 is given by @xmath70 \;.\end{aligned}\ ] ] the functions @xmath71 , @xmath72 and @xmath73 are defined as @xmath74 @xmath73 is the coulomb term and @xmath75 is the chern - simons term of the rpa formula .
the second and third summands @xmath73 and @xmath76 in ( [ 5105 ] ) are given by the first order graphs of the rpa which have to be subtracted for treating equal time green s function in the right way @xcite .
the density - density response function @xmath77 and the transversal current - current response function @xmath78 are given in ( [ 30280 ] ) of appendix a. the hartree - fock terms of @xmath63 are given by @xmath79 the function @xmath80 is the fermion occupation factor for the chemical potential @xmath81 .
the vertices @xmath82 are defined by @xmath83 here @xmath84 is the coulomb vertex .
@xmath85 is the chern - simons vertex @xmath86 mentioned in the introduction .
comparing @xmath87 with the analogous expression in @xcite the only difference between the two expressions is the hartree term @xmath68 in ( [ 5135 ] ) , which corresponds to a hartree feynman graph of @xmath59 ( [ 1050 ] ) resulting from the normal ordering of the @xmath88 term .
it is shown in @xcite , that @xmath89 is ir - divergent .
this is based on the ir - divergence in @xmath90 . the additive term @xmath68 shifts this ir - divergent grand canonical potential in @xcite to an uv - divergent grand canonical potential @xmath87 . as mentioned above @xmath90 and @xmath68
belong to the hartree - fock feynman graphs of @xmath91 .
therefore , if we calculate the energy of all hartree - fock feynman graphs in addition to @xmath66 we will get a finite energy because the hartree - fock energy of @xmath59 ( [ 1050 ] ) has to be finite .
furthermore , it is easily seen that @xmath66 corresponds to the maximal divergent graphs for a given number of vertices . to get a good approximation of the energy we take into account in the following the maximal divergent graphs ( per number of vertices ) together with the hartree - fock graphs . as mentioned in the introduction this energy principle
is also used for a calculation of the ground state energy of the coulomb gas by many authors ( e.g. @xcite ) . in order to calculate that energy
we have to determine at first the hartree - fock energy of @xmath59 .
sitko calculated in @xcite the hartree - fock energy of the magnetic part of @xmath59 .
the coulomb part of the hartree - fock energy consists only of @xmath92 ( [ 5140 ] ) .
this integral is easily evaluated yielding @xmath93 for the hartree - fock energy of the chern - simons hamiltonian @xmath94 . the first term in ( [ 2315 ] )
is the kinetic energy of the composite fermions . for the @xmath0 system we have to insert @xmath18 into @xmath95 .
doing this we obtain two times the value of the exact magnetic energy for @xmath95 ( @xmath96 ) .
+ with the help of an @xmath97-expansion of @xmath66 in ( [ 5105 ] ) we can calculate the coefficients of the expansion for the @xmath14 system numerically : @xmath98 the elctron density @xmath99 is given by @xmath100 .
this can be motivated by the luttinger - ward theorem @xcite under the consideration that there are no anomalous graphs in our approximation of the energy .
thus we get from ( [ 2315 ] ) , ( [ 2320 ] ) @xmath101 for the energy density @xmath102 of the @xmath0 system
. it is evident by the selection principle of the calculated feynman graphs that the coulomb energy part in @xmath103 is the same as in @xcite . to compare our perturbative result for the energy with the energy obtained by other methods we have written down in @xmath104 the exact magnetic energy and the coulomb energy obtained by numerical diagonalization techniques
this coulomb energy was calculated by morf and dambrumenil @xcite as well as girlich @xcite by diagonalization of the coulomb part of the hamiltonian for electrons on a sphere in the lowest landau level .
one sees that perturbational and the numerical calculated energies are in satisfactory agreement .
in the @xmath0 as in the @xmath3 system appears an interesting phase transition from a spin unpolarized ground state to a spin polarized ground state by modifying the interaction strength between the electrons . in the next few sections we will see , that this physical phenomenon already exists on the level of the rpa of the chern - simons theory . to show this we have to add a spin constraint variable into the chern - simons path integral . without this spin - constraint variable
the @xmath0 chern - simons path integral is written as in ( [ 1150 ] ) with a doubling of the fermionic grassmann fields representing the spin - up and spin - down freedom of the electrons . in the following
we denote by @xmath105 , ( @xmath106 ) the creation operator of one composite fermion at point @xmath107 with spin up ( spin down ) .
@xmath108 , ( @xmath109 ) annihilates one composite fermion with spin up ( spin down ) at point @xmath110 .
+ the total spin in the xy - plane is given by @xmath111 furthermore , we define from @xmath112 ( [ 1050 ] ) a new spin constraint chern - simons hamiltonian @xmath113 by @xmath114 .
we will fix the total spin @xmath115 by the constraint such that @xmath116 is minimal for a given @xmath117 .
this is the case if @xmath118 .
one can get this constraint on @xmath116 by differentiating @xmath119 with respect to @xmath120 .
thereby we denote the partition function of @xmath121 by @xmath122 .
thus we get the following conditions on the partition function @xmath123 @xmath124 the symbol @xmath125 is the expectation value in the gibb s state of the hamiltonian @xmath126 .
@xmath127 is the particle number operator @xmath128 .
@xmath129 and @xmath130 are the chemical potentials of the system representing the spin - up and spin - down electrons .
furthermore , we have to mention that one has to be careful in using equation ( [ 30145 ] ) .
this is because the first term on the right hand side of equation ( [ 30145 ] ) equals to @xmath131 for a finite system .
this is no longer the case for an infinite system ( an example is the kinetic hamiltonian ) .
thus one has to differentiate first for a finite system and afterwards take the limit to an infinite system .
we are now able to write down the @xmath0 chern - simons path integral . using the hubbard - stratonovich decoupling of the fermionic fields in the operator @xmath132 we get @xmath133 { \cal d}[c^*_{\kappa , l},c_{\kappa , l } ] { \label{30150 } } \\ & & \hspace{0.2cm}\times \exp\left[-\epsilon\left ( l^s_{l}+l_{s , l}+l_{ee , l}+l^0_{ee , l } + l_{cs , l}+l^0_{s , l}\right)\right ] \nonumber \end{aligned}\ ] ] with the help of @xmath134 and the normalizing factor @xmath135 \nonumber \\ & & \hspace{1 cm } \times \exp\left[-\epsilon\left(l_{cs , l}+l^0_{ee , l}+l^0_{s , l}\right)\right ] { \label{30170}}\;.\end{aligned}\ ] ] the terms @xmath136 , @xmath137 in ( [ 30150 ] )
are given by the hubbard - stratonovich decoupling of the spin term @xmath132 .
after integrating in ( [ 30150 ] ) over the fermionic fields we get the following mean - field equations for the bosonic fields @xmath138 the mean - field green s function is given by @xmath139 .
integrating this mean - field expansion to second order of the bosonic fields we get for the grand canonical potential @xmath140 .
@xmath141 is the grand canonical potential for a spin dependent electron gas without interaction . as we do not have any anomalous graphs , we want to make use of the luttinger - ward theorem @xcite .
this theorem states that a good approximation is obtained for @xmath129 , @xmath142 if one uses @xmath143 and @xmath144 .
@xmath145 is the area of the system . as a function of @xmath129 , @xmath142
, the expectation value @xmath146 behaves like a step function for a finite system .
because the first term on the right hand side of equation ( [ 30145 ] ) is @xmath147 we observe that this term is zero . because @xmath62 does not depend on @xmath120 , it can not fulfill the spin constraint equation ( [ 30145 ] ) . on the other hand
it is clear that the ground state of the mean - field hamiltonian fulfills the spin constraint @xmath148 . for
this reason @xmath62 together with the grand canonical potential of the exchange graph of @xmath149 ( [ 30110 ] ) fulfills the spin constraint equation ( [ 30145 ] ) ( the exchange graph is a part of @xmath150 ) . furthermore
, we have to mention that the exchange part of @xmath149 is the only part of @xmath150 which is linear in @xmath120 .
thus the spin constraint equation ( [ 30145 ] ) is fulfilled for @xmath151 . in the following
we denote @xmath152 by @xmath153 .
@xmath154 are the diverse response functions of appendix a. after some transformations and an expansion in @xmath155 we get with the help of the functions @xmath156 the @xmath97-part of the grand - canonical @xmath157 as @xmath158 after developing the chern - simons theory of the @xmath3 system in the next section , we will calculate @xmath159 in section v numerically .
the @xmath3 system consists of one lowest landau level filled by spin up and spin down electrons .
the second landau level is half filled . after performing the chern - simons transformation of the hamiltonian of this system one
does not get a mean field free theory as in the case of the @xmath0 system .
it is clear that the filled lowest landau level of the @xmath3 system should not have much relevance for the physical properties .
thus we will consider the @xmath3 system as a @xmath0 system with a modified interaction potential between the electrons .
if we define an isometric transformation which transforms a wave function of a higher landau level to the next lower landau level then the calculation of the coulomb energy in the second landau level corresponds to the calculation of the energy of the first landau level with the transformed interaction potential of the electrons . if one calculates the coulomb energy by perturbational methods then this transformation should not change the wave function too much .
we will develop in the following an isometric transformation which leaves the center of mass of the wave function invariant .
+ the ladder operators between landau levels are defined with the help of the operators @xmath160 and @xmath161 by @xmath162 the ladder operators act on a wave function belonging to the n - th landau level as follows @xmath163 with the help of some commutation relations @xcite one can easily see that @xmath164 the same relation holds for the @xmath165-coordinate .
the partial isometric transformation @xmath166 which is descending the landau level functions is given by @xmath167 with the help of the ladder operators ( [ 30010 ] ) the operator @xmath166 is written by @xmath168 it is easily seen from equation ( [ 30080 ] ) that the n - th term in @xmath169 transforms a wave function which belongs to the n - th lowest landau level to zero .
the higher order terms in @xmath169 are motivated by the different normalization factors of the individual landau levels in equation ( [ 30030 ] ) .
we observe from the equations ( [ 30030 ] ) and ( [ 30035 ] ) that the operator @xmath169 leaves the center of mass coordinate of one landau level function invariant .
+ we will use this operator @xmath169 for transforming the coulomb interaction to one lower landau level .
almost all publications carrying out ground state energy calculations of the @xmath1 and the @xmath3 system by numerical methods do not take into account landau level splitting @xcite . in these calculations
one diagonalizes the coulomb interaction operator in the second landau level in the case of the @xmath3 system .
it is easily seen from perturbation theory that this energy is given by the @xmath155-term of the coulomb energy . because we want to compare the energy given by perturbational calculations and numerical calculations we can neglect the higher order @xmath170 terms in @xmath169 under the condition of calculating the coulomb energy to order @xmath171 .
so we get for the isometric partially transformed coulomb interaction operator @xmath172 @xmath15 is given by @xmath173 where @xmath99 is the density of the electrons in the second landau level .
since one does not start from a chern - simons transformed wave function in the lowest landau level in the perturbational calculation but from a slater determinant of plane waves , one should also take into account higher order terms of @xmath169 in perturbation theory . according to equation ( [ 30030 ] )
these only differ from the first order term by a normalizing factor .
therefore we will fix an effective @xmath15-field @xmath174 in equation ( [ 30090 ] ) when proceeding @xmath175 this effective field will be brought into line with property ( [ 30080 ] ) .
the chern - simons transformed operator @xmath176 of @xmath169 is constructed from @xmath169 in ( [ 30080 ] ) by the substitution @xmath177 .
it is approximated through @xmath178 with the operator @xmath179 constructed from @xmath180 by the substitution @xmath177 .
the @xmath3 chern - simons path integral @xmath181 is constructed from the @xmath0 chern - simons path integral @xmath182 by the substitution @xmath183 .
the function @xmath184 is given by @xmath185 the background field @xmath29 is given by @xmath186 @xmath187 is constructed from @xmath169 in ( [ 30080 ] ) by the substitution @xmath188 .
because of this equation the hartree graphs in the @xmath3 chern - simons theory are cancelled by the @xmath29 couplings . like in the @xmath0 system
we are able to calculate the coulomb energy of the path integral @xmath189 in rpa . like
in the case of the @xmath0 system we obtain @xmath190 with the inclusion of the spin constraint ( [ 30145 ] ) .
we now calculate the coulomb energy of the path integral ( [ 30150 ] ) , ( [ 30093 ] ) , with the approximation @xmath191 in rpa .
this approximation is also used in ( [ 30096 ] ) . for this
we define the @xmath3 coulomb potential by @xmath192 .
with the help of the expressions @xmath193 we get the rpa coulomb energy by @xmath194 the several response functions @xmath154 in ( [ 30340 ] ) are defined in ( [ 30270 ] ) of the appendix .
now we have to fix the effective magnetic field @xmath195 by perturbational methods .
we will treat at first the case of a spin polarized ground state ( we have to insert @xmath196 and @xmath197 in ( [ 30340 ] ) and ( [ 30320 ] ) )
. for fixing @xmath195 we have look for the main contributions to the momentum integrals of @xmath198 in ( [ 30345 ] ) .
we will fix @xmath195 so that we get for these momentums the same integrand as we would calculate the path integral ( [ 30150 ] ) , ( [ 30093 ] ) with the exact operator @xmath199 ( [ 30080 ] ) in @xmath200 .
numerical integration of the two terms in @xmath201 ( [ 30345 ] ) yields results for @xmath202 which are about a factor 3 smaller compared to the second term .
this is in contrast to the @xmath0 system . when fixing @xmath203 in @xmath204 there is no impact on the proportion of the two terms .
thus we use the second term in ( [ 30345 ] ) for fixing @xmath205 . after performing the @xmath206 integration in ( [ 30345 ] ) the @xmath207 integrand of the second term
has the largest contribution to the energy for @xmath208 .
this is shown in figure [ param52 ] .
thus the ring momentums @xmath209 are the most important wave vectors in the calculation of the response functions ( [ 30270 ] ) in order to calculate the ground state energy of the @xmath3 system .
this is due to a @xmath210 term as a result of integrating ( [ 30270 ] ) over the ring frequencies .
thus we find the following equation to get the same @xmath211-integrand in the second term in ( [ 30345 ] ) using the exact @xmath176 ( [ 30080 ] ) and its approximation ( [ 30091 ] ) in the path integral @xmath212 @xmath213 is the one particle function @xmath214 $ ] where @xmath215 is a vector of length @xmath216 .
furthermore , we have to insert the mean - field condition @xmath217 in @xmath218 of equation ( [ 30390 ] ) . by using the definitions ( [ 30038 ] ) , ( [ 30080 ] ) the left hand side of equation ( [ 30390 ] ) is equal to @xmath219 in this equation @xmath169 is calculated with the vector potential @xmath220 .
the magnetic field @xmath221 is defined by @xmath222 . from the equations ( [ 30390 ] ) and ( [ 30391 ] ) we obtain @xmath223 .
+ in the case of a spin unpolarized ground state we get with the help of @xmath224 the approximation @xmath223 due to the validity of the equations ( [ 30390 ] ) , ( [ 30391 ] ) . in the case of @xmath225
we obtain the equation @xmath226 to get the same @xmath211-integrand in the second term of equation ( [ 30345 ] ) using the exact @xmath176 ( [ 30080 ] ) and its approximation ( [ 30091 ] ) in the path integral ( [ 30150 ] ) , ( [ 30093 ] ) .
if we compare this equation for @xmath196 with ( [ 30390 ] ) and ( [ 30391 ] ) we get a factor two difference .
this is because we used for the response functions @xmath227 in equations ( [ 30390 ] ) and ( [ 30391 ] ) .
this is only correct for @xmath228 ( see e.g. @xmath229 in ( [ 30280 ] ) ) .
one has to notice that @xmath230 is non - zero only on a non - measurable subspace of the @xmath231-integral of the second term of @xmath198 ( [ 30345 ] ) .
thus @xmath230 is not relevant for the energy calculation and @xmath223 is a good approximation for @xmath232 .
the absolute value @xmath233 of the response functions ( with any vertex functions ) are growing functions of @xmath234 for @xmath235 and zero for @xmath236 and @xmath235 . therefore ,
for a determination of @xmath205 we only consider that part of the @xmath211 reponse functions of the second term of @xmath198 ( [ 30345 ] ) which contains the larger @xmath237-value @xmath238 of @xmath129 , @xmath142 .
the demand that this part of the @xmath211-integrand is identical by using in the path integral either the exact @xmath176 ( [ 30080 ] ) or its approximation ( [ 30091 ] ) , results in @xmath239 .
thus we get for @xmath240 of the spin polarized as well as the spin unpolarized case the former values .
+ in the following we calculate the compressibility , the effective mass and the excitations of the spin polarized @xmath4 systems .
this is done by a restriction of the path integral ( [ 30150 ] ) , ( [ 30093 ] ) to one spin component ( and inserting @xmath190 ) .
we now use standard techniques @xcite to calculate the compressibility , the effective mass and the excitations of the @xmath4 systems in rpa . to calculate the compressibility we have to pay attention to the construction of the path integral of the @xmath4 systems .
it is then easy to see that in the above model of the spin polarized @xmath4 systems the density - density correlation function is given by @xmath241 . in this expression @xmath242
is given by the quantum mechanical heissenberg operator of the expression @xmath243 .
the fourier transformed density - density correlation function could be calculated by analytical continuation of the fourier transformed sum of @xmath244 and @xmath245 .
@xmath244 is the @xmath246 correlation function of the path integral ( [ 30150 ] ) , ( [ 30093 ] ) . with the help of the approximations ( [ 30430 ] )
we obtain for the ( retarded ) density - density correlation function @xcite of the @xmath4 system in the range @xmath247 and @xmath248 @xmath249 here @xmath250 is the time ordering operator . from ( [ 30500 ] ) we see that @xmath251 is zero for the @xmath3 system . by considering higher order terms of the response functions ( [ 30280 ] ) in the momentum @xmath252 and @xmath253
we get for the @xmath3 system @xmath254 . with the help of the retarded density - density correlation function it is possible two
calculate the compressibility @xmath255 of the system by using the compressibility sum rule @xcite @xmath256 .
due to this relation we obtain by the definition of @xmath195 that the compressibility is given by @xmath257 .
thus we see that the systems of filling fraction @xmath4 ( @xmath258 ) should be compressible .
furthermore we get the same rpa compressibility as hlr calculated for the @xmath0 system except for a factor @xmath259 . in the case of filling fraction @xmath3
this is no longer valid .
due to our calculation the @xmath3 system is incompressible . finally , if we take into account all terms of @xmath169 ( [ 30080 ] ) in the rpa calculation of the compressibility , we get the same expression as ( [ 30500 ] ) for the @xmath260 systems .
this is due to the definition of @xmath205 ( [ 30390 ] ) .
+ next we calculate the effective mass @xmath261 in rpa . here
@xmath262 is the self energy of the fermions of the path integral ( [ 30150 ] ) , ( [ 30093 ] ) . due to the same singular structure of the transversal ( @xmath263 ) propagator we obtain for the @xmath264 system with the help of ( [ 30430 ] ) a singular effective mass which is different by a factor @xmath265 from the ( diverging ) effective mass of the spin polarized @xmath0 system calculated by hlr .
the expression for the compressibility as well as the effective mass suggests that systems of filling fraction @xmath4 behave similar to the @xmath0 system except @xmath266 .
this is in agreement with numerical calculations by morf and dambrumenil @xcite . + like in the @xmath0 system one
can also derive the cyclotron excitations for the @xmath4 systems by a calculation of the density - density correlation function . in this calculation
one has to take into account the response functions in the limit ( [ 30510 ] ) .
we have calculated the integrals in ( [ 30370 ] ) , ( [ 30345 ] ) using numerical methods .
the results are shown in figure [ bild1 ] .
this figure shows the coulomb energy as a function of the strength @xmath267 of the coulomb interaction function @xmath268 .
it contains the calculation of the spin polarized system , i.e. @xmath269 , @xmath270 , as well as the spin unpolarized system , i.e. @xmath271 , for the @xmath0 and the @xmath3 system .
the graph reveals a transition from a spin singlet to a spin polarized ground state at @xmath272 for the @xmath0 system . in the @xmath3 system
the transition is at @xmath273 .
this coincides qualitatively with the numerical result of jain et al .
@xcite , which has be obtained by exact diagonalization methods on a sphere . since jain et al . employed the haldanes pseudo potentials @xmath274 , where the strength of the coulomb interaction is adjusted on the basis of the ratio of @xmath275 to @xmath276 we can not quantitatively compare the two results @xcite .
that is , jain et al .
modify the strength of the pseudo potentials with the help of @xmath277 in @xmath278 $ ] . for the coulomb interacting @xmath1 system
the result is @xmath279 . for the coulomb interacting @xmath12 system
one gets @xmath280 .
jain et al .
obtained a transition from a spin singlet to a spin polarized ground state for the @xmath1 system at @xmath281 .
they obtained a good overlap of the ground state wave function with a spin polarized chern - simons wave function for @xmath282 .
the same transition was observed at @xmath283 for the @xmath12 system . because @xmath284 is dominated by the strength of the interaction of neighbouring electrons , the transition at @xmath285 showing in figure ( [ bild1 ] )
agrees with the results of jain et al .. but comparing the value @xmath267 at the transition of the @xmath0 system with the @xmath3 system , @xmath267 of the @xmath3 system is probably too small .
furthermore , the slope of the two curves in figure [ bild1 ] of spin @xmath286 ( @xmath287 with @xmath288 ) and @xmath289 of the @xmath3 system differs only slightly .
thus the @xmath3 system compared to the @xmath0 system can change the alpha - value of the transition very easily due to the higher orders of perturbation theory .
+ morf shows in @xcite using similar numerical diagonalization methods like jain et al . but considering a larger number of electrons that the ground state of the @xmath3 system is not a chern - simons wave function for @xmath290 ( coulomb potential ) .
this ground state wave function shows a large overlap with a spin polarized incompressible paired wave function . for smaller @xmath291 , respectively @xmath267
, morf could not identify the ground state wave function .
recently rezayi and haldane @xcite could identify this ground state as a compressible stripped phase by using similar numerical diagonalization methods like morf . for larger @xmath291 , respectively @xmath267 ,
morf @xcite obtains for the ground state wave function a large overlap with a spin polarized chern - simons wave function .
based on this observation he interpreted the experiments of eisenstein et al.@xcite , mentioned in the introduction , that caused on the tilted magnetic field the effective interaction between two neighbouring electrons is increased .
thus the ground state wave function of the system changes to a spin polarized chern - simons wave function .
at last we mention that rezayi and haldane get in @xcite a transition of a compressible stripped phase via an incompressible paired quantum hall state to a chern - simons ground state near @xmath280 even though they considered the pure spin polarized @xmath3 system .
this shows that the transition is an interaction effect ( not caused by the spin degree of freedom ) .
+ comparing for @xmath290 the coulomb energies @xmath292 , @xmath293 of the ground state ( spin polarized ) with the coulomb energies @xmath294 , @xmath295 of the numerical calculation method @xcite we obtain @xmath296 figure [ bild2 ] shows the coulomb energy as a function of the variable @xmath297 for @xmath290 of the @xmath1 and the @xmath12 system .
@xmath298 , @xmath299 are given by @xmath300 .
the total spin @xmath301 is determined by @xmath302 .
the figure shows that the spin polarized ground state , i.e. @xmath303 , has a minimal coulomb energy of all @xmath116 ground states .
this is in contradiction with recent experimental obeservations @xcite which shows that the @xmath0 ground state is not fully polarized .
unfortunately these experiments do not show the degree of polarization of the ground state .
we showed that due to a wrong operator ordering in the path integral of the spin polarized @xmath0 system of hlr we got in @xcite an ir divergence in the ground state magnetic energy .
therefore we used the well - known chern - simons partition function in an operator formulation to get a chern - simons path integral with a correct operator ordering .
we calculated the energy of this path integral in rpa to show that this path integral has a uv infinite magnetic energy .
this uv infinity can be corrected by considering the complete set of hartree - fock feynman graphs of the chern - simons hamiltonian . when calculating the energy of the maximal divergent graphs together with the hartree - fock graphs we obtained a well behaved finite magnetic energy .
the coulomb energy is the same as in @xcite .
+ next we developed a formalism to calculate the coulomb ground state energy of the @xmath0 system subjected to a spin @xmath116 constraint and calculated the coulomb energy as a function of the interaction strength @xmath304 .
this was done for a spin polarized ground state as well as for a spin unpolarized ground state .
furthermore , we formulated a chern - simons theory of the @xmath4 systems by transforming the interaction operator of the electrons from the second to the lowest landau level getting a @xmath305 chern - simons theory with a modified interaction operator .
we used this chern - simons theory to calculate the compressibility , the effective mass and the excitations of the spin polarized @xmath4 systems in rpa .
we see from our calculations that @xmath3 is a special system within all @xmath4 systems .
we get the same compressibility for the @xmath4 systems as hlr calculated for the @xmath0 system except for a factor @xmath306 .
thus we get that the @xmath4 systems are compressible except for @xmath307 . due to our rpa calculation
the @xmath3 is incompressible .
next we calculated the effective mass of the @xmath4 systems .
for these systems we get a diverging effective mass like hlr got for the @xmath0 system with a difference by a factor of @xmath308 ( @xmath309 ) .
the effective mass of the @xmath3 system is the same as hlr calculated for the @xmath0 system without interaction between the electrons .
these calculations ( compressibility and effective mass ) are in agreement with numerical calculations by morf and dambrumenil @xcite which show that systems of filling fraction @xmath4 behave similar to the @xmath0 system except for @xmath310 . like in the calculation of the cyclotron excitations of the @xmath0 system by hlr
we also obtain the cyclotron excitations for the @xmath4 systems in rpa .
+ next we used the @xmath3 chern - simons theory to calculate the coulomb energy as a function of the interaction strength for the spin polarized as well as the spin unpolarized ground state in rpa .
we get for the @xmath0 as well as the @xmath3 system a transition from a spin unpolarized ground state to a polarized ground state at @xmath311 .
this agrees with calculations using numerical diagonalization techniques .
furthermore , for @xmath312 we get a good correspondence to the energies calculated by these numerical methods .
at last we calculated the ground state coulomb energies for the @xmath0 and the @xmath3 system as a function of the total spin @xmath116 for @xmath313 .
we saw that for both systems the maximal spin polarized ground states is actually in the coulomb energy minimum of all @xmath301 states which is not in agreement with recent experimental observations @xcite .
+ summarizing , already on the level of the rpa , we can see spin effects for both the @xmath0 and the @xmath3 systems which agree with numerical simulations .
as mentioned earlier it is supposed that the coulomb @xmath3 chern - simons ground state wave function ( @xmath314 ) is of a pfaffian type .
thus the goal still is to develop a theory of the coulomb interacting @xmath3 system which has a mean - field ground state of the paired wave type .
we would like to thank m. hellmund , k. luig and w. weller for many helpful discussions during the course of this work .
further we have to acknowledge the financial support by the deutsche forschungsgemeinschaft , graduiertenkolleg `` quantenfeldtheorie '' .
in this section we calculate the response functions for the @xmath0 and the @xmath3 system .
these response functions were used in the energy formulas ( [ 5105 ] ) , ( [ 30370 ] ) , ( [ 30345 ] ) .
these response functions are defined by @xmath315 g is the green s function @xmath316 . after performing the integrals in ( [ 30270 ] ) one gets with the help of the substitution @xmath317 @xmath318 \;,\nonumber \\ & & \pi_{\sigma \sigma}^2(q,\omega,\mu ) = \frac{m^3}{4 \pi } \sum_{\sigma \in \{-1,1\}}\left [ 2\frac{(i\sigma\omega)}{3u^2 } \bigg\{(u+i\sigma\omega)^3 \right .
\nonumber \\ & & \left.-\sqrt{(u+i\sigma\omega)^2 - 4u\mu } \left((u+i\sigma\omega)^2 + 2u\mu\right)\bigg\}-4\mu^2 \right ] \ ; , \nonumber \\ & & \pi_{\sigma a^0}^1(q,\omega,\mu ) = \frac{m^2}{4 \pi } \sum_{\sigma \in \{-1,1\}}\left[-\frac{1}{3 u^2 } ( u+i\sigma\omega)^3 \right .
\nonumber \\ & & \hspace{0.8 cm } \left.+\frac{1}{3 u^2}\sqrt{(u+i\sigma\omega)^2 - 4u\mu } \left((u+i\sigma\omega)^2 + 2u\mu\right ) \right ] \ ; , \nonumber \\ & & \pi_{\sigma a^0}^2(q,\omega,\mu ) = \frac{m^2}{4 \pi } \sum_{\sigma \in \{-1,1\}}\bigg[\frac{(u+i\sigma\omega)}{u}\nonumber \\ & & \hspace{1.3 cm } \times \left((u+i\sigma\omega ) -\sqrt{(u+i\sigma \omega)^2 - 4u\mu } \right)-2\mu \bigg]\ ; , \nonumber \\ & & \pi_{\sigma a}(q,\omega,\mu ) = -i\ , m^{\frac{3}{2}}\sqrt{2u}\;\pi_{aa}(q,\omega,\mu ) \;,\nonumber \\ & & \pi_{aa}(q,\omega,\mu ) = \frac{1}{2\pi}{\tmspace{-}{2\thinmuskip}{2\txtmu } } \sum_{\sigma \in \{-1,1\}}{\tmspace{-}{2\thinmuskip}{2\txtmu } } \bigg [ \frac{1}{12 u^2}\bigg\ { \left ( u+i \sigma \omega \right)^3 \nonumber \\ & & \hspace*{2.5 cm } -\sqrt{\left(\left(u+i \sigma \omega \right)^2- 4 u \mu\right)^3 } \bigg\}-\frac{\mu}{2}\bigg ] , \nonumber \\ & & \pi_{a^0 a^0}(q,\omega,\mu ) = \frac{m}{2\pi}{\tmspace{-}{3\thinmuskip}{3\txtmu } } \sum_{\sigma \in \{-1,1\}}{\tmspace{-}{3\thinmuskip}{3\txtmu } } \left[\frac{1}{2u } \sqrt{\left(u+i \sigma \omega \right)^2 - 4 u \mu}-\frac{1}{2 } \right]{\tmspace{-}{2\thinmuskip}{2\txtmu}}.\nonumber \end{aligned}\ ] ] @xmath319 and @xmath320 are given by @xmath321 for calculating the compressibility , the effective mass and the excitations we have to continue analytically the response functions @xmath154 to imaginary frequencies @xcite .
in the range @xmath322 and @xmath323 we get by multiplying every value in ( [ 30430 ] ) by @xmath325 , one can calculate the asymptotic of the next term in the expansion of the response functions .
comparing @xmath326 in ( [ 30430 ] ) with the corresponding term of hlr @xcite , the result differs by a factor two in the first term of @xmath327 .
this difference is also observed in the case of the @xmath0 chern - simons system with impurities @xcite .
+ furthermore , we need the analytical continued response functions in the range @xmath328 and @xmath329 for a calculation of the cyclotron excitations . in this range
we get uwe girlich , ph .
d. thesis , university of leipzig ( 1999 ) ( unpublished ) .
* 61 * , 997 ( 1988 ) ; j.p .
et al . _ ,
sci . * 229 * , 31 ( 1990 ) ; j.p .
et al . _ ,
* 62 * , 1540 ( 1989 ) . | we reconsider energy calculations of the spin polarized @xmath0 chern - simons theory .
we show that one has to be careful in the definition of the chern - simons path integral in order to avoid an ir divergent magnetic ground state energy in rpa as in @xcite .
we correct the path integral and get a well behaved magnetic energy by considering the energy of the maximal divergent graphs as well as the hartree - fock graphs .
furthermore , we consider the @xmath1 and the @xmath2 system with spin degrees of freedom . in
doing this we formulate a chern - simons theory of the @xmath3 system by transforming the interaction operator to the next lower landau level .
we calculate the coulomb energy of the spin polarized as well as the spin unpolarized @xmath1 and the @xmath2 system as a function of the interaction strength in rpa .
these energies are in good agreement with numerical simulations of interacting electrons in the first as well as in the second landau level .
furthermore , we calculate the compressibility , the effective mass and the excitations of the spin polarized @xmath4 systems where @xmath5 is an even number . 2 |
graphene , a single atomic layer of graphite , has been successfully produced in experiment @xcite , which has resulted in intensive investigations on graphene - based structures , due to the fundamental physics interests that is involved and the promising applications @xcite .
there are significant current efforts devoted to growing graphene epitaxially @xcite by thermal decomposition of silicon carbide ( sic ) , or by vapor deposition of hydrocarbons on catalytic metallic surfaces , which could later be etched away , leaving graphene on an insulating substrate .
the low energy quasiparticle excitations in graphene are linearly dispersing , and are described by dirac cones at the edges of the first brillouin zone .
the linear energy - momentum dispersion has been confirmed by recent observations @xcite .
the slope of the linear relation corresponds the fermi velocity of chiral dirac electrons in graphene , which plays an essential role in the landau - fermi liquid theory @xcite and has a direct connection to the experimental measurement .
there are some unusual features of graphene , such as the effects of electron - electron interactions on the ground - state properties @xcite , anomalous tunneling effect described by the klein tunneling , the tunneling through a p - n junction@xcite that follows from chiral band states , and the energy - momentum linear dispersion relation .
the klein tunneling predicts that the chiral massless carrier can pass through a high electrostatic potential barrier with probability one , regardless of the height and width of the barrier at normal incidence , which is in contrast with the conventional nonrelativistic massive carrier tunneling where the transmission probability decays exponentially with the increasing of the barrier hight and would depend on the profile of the barrier .
@xcite an exciting experimental development is the ability to apply an electric field effect or submicron gate voltage , in order to illustrate graphene p - n junctions .
@xcite by applying an external gate voltage , the system can be switched from the n - type to the p - type carriers , thereby controlling the electronic properties that give rise to graphene - based nanodevices .
recently , strong evidence for klein tunneling across potential steps which is steep enough in graphene has been experimentally observed .
@xcite clean graphene junctions were predicted to display a number of fascinating physical phenomena , even in the absence of electron - electron interactions .
@xcite interestingly , the veselago lensing of electric current by a single p - n junction in clean graphene @xcite and the andreev reflection , and the electron to hole conversion at the interface at normal incidence @xcite , have all been predicted theoretically .
such phenomena are predicted to change both quantitatively and qualitatively when disorder is included in the model .
for instance , inhomogeneous graphene p - n junction systems were studied using the thomas - fermi approximation , including disorder effects , by fogler and collaborators @xcite .
they showed that junction resistance is dominated by either ballistic or diffusive contributions , depending on the density of charged impurity and gradient of the carrier density . in the semiconductor context
there are basically a large number of works on the tunneling , which have resulted in the `` obvious '' declaration that the electronic properties of semiconductor superlattice are different from those calculated in a single - barrier junction .
moreover , the electronic properties of semiconductor superlattices in the presence of disorder have been studied by several groups @xcite .
importantly , all the electronic states are localized in the thermodynamic limit for a semiconductor superlattice in the presence of white - noise disorder .
@xcite graphene superlattices , on the other hand , may be fabricated by adsorbing adatoms on graphene surface through similar techniques , by positioning and aligning impurities with scanning tunneling microscopy @xcite or by applying a local top gate voltage to graphene .
@xcite recently , a periodic pattern in the scanning tunneling microscope image has been demonstrated on a graphene on top of a metallic substrate .
@xcite the transition of hitting massless particles in graphene - based superlattice structure ( gsls ) was first studied by bai and zhang .
@xcite they showed that the conductivity of the gsls depends on the superlattice structural parameters .
furthermore , the superlattice structure of graphene nanoribbons has been recently studied by using first - principles density functional theory calculations .
@xcite these calculations showed that the magnetic ground state of the constituent ribbons , the symmetry of the junction , and their functionalization by adatoms represent structural parameters to the electronic and magnetic properties of such structures .
recently , novel physical properties of gsls with one - dimensional ( 1d ) kronig - penney type and 2d muffin - tin type potentials were also studied .
@xcite the results showed that a periodic potential applied by suitably patterned modifications leads to further charge carrier behavior .
the propagation of charge carriers through such a superlattice is highly anisotropic , and in extreme cases results in group velocities that are reduced to zero in one direction but are unchanged in the other direction .
moreover , they showed that the density and type of carrier are extremely sensitive to the applied potential .
it would , therefore , be worthwhile to investigate how the conductance of graphene superlattice junctions are affected by structural white noise , and compare the conductances with those calculated for disordered semiconductor superlattice . due to the conservation of pseudospins in graphene ,
backscattering process is suppressed at normal incidence , which makes the disordered regions transparent .
@xcite the purpose of this paper is to study the electronic behavior of graphene superlattices p - n junctions by using the transfer - matrix method .
the system that we study consists of a sequence of electron - doped graphene as wells , and hole - doped graphene as barriers .
we study the effect of the disorder imposed on the size of the barriers in the transmission probability , @xmath0 , through the system as a function of the system size ( number of the barriers ) , together with various incident angles .
the dc conductance of the finite - size system takes on a nonzero value of the transmission in some special configurations . using the finite - size scaling of transmission
, we show that the conductance , in the thermodynamic limit , tends to a finite constant for spacial cases .
the rest of this paper is organized as follows . in sec .
ii we introduce the models and derive the related transfer matrix .
we also explain how we calculate the transmission probability and the dc conductivity .
section iii contains our numerical calculations .
we conclude in sec .
iv with a brief summary .
consider a system of superlattice p - n junctions in the independent carriers model at zero temperature , and in the absence of carrier - phonon and spin - orbit interactions .
the low - energy massless dirac - band hamiltonian of graphene in the continuum model can be written as@xcite @xmath1 where @xmath2 for the inequivalent @xmath3 and @xmath4 valleys at which @xmath5 and @xmath6 bands touch , @xmath7 is an envelope function momentum operator , @xmath8 is the fermi velocity , and @xmath9 is a pauli matrix that acts on the sublattice pseudospin degree of freedom .
the total hamiltonian of a massless carrier in a special geometry is written as , @xmath10 where @xmath11 is the graphene - based superlattice potential which is modeled as described below .
we consider superlattice p - n junctions in a graphene - based structure .
the system consists of two kinds of graphene with different potentials , the first being an electron - doped graphene with thickness @xmath12 , while the second is a hole - doped part with thickness @xmath13 , standing alternately .
the potential for the electron- and hole - doped graphene are @xmath14 and zero , respectively .
the energy of the incident particle is @xmath15 with the wavelength @xmath16 across the barriers , in such a way that the fermi level lies in the conduction band outside the barrier and the valence band inside it , i.e. , @xmath17 , as shown in fig .
the growth direction is taken to be the @xmath18 axis which is designed as the superlattice axis . in order to neglect the strip edges , we assume that the width of the graphene strip is much larger than @xmath13 .
we set disorder situations in which the value of @xmath13 fluctuates around its mean value , given by @xmath19 . in the model
the fluctuations are given by , @xmath20 , where @xmath21 is a set of uncorrelated random variables or white noise with the box distribution , @xmath22 , and @xmath23 is the site index . here , the @xmath24 is the disorder strength .
we consider graphene - based superlattice potential in a simple model as @xmath25 where @xmath26 is the position of barriers center .
the model is similar to the potential of semiconductor superlattices that has been used by other groups .
@xcite let us now consider the case in which the incident massless electron in the gsls propagates at angle @xmath27 along the @xmath18 axis ( see fig
. 1 ) and , therefore , the dirac spinor components , @xmath28 and @xmath29 , which are the solutions to the dirac hamiltonian , can be expressed @xcite as : @xmath30 where @xmath31 and @xmath32 in order to calculate the transmission coefficients , we use the transfer - matrix method @xcite . to this end
, we apply the continuity of the wave function at the boundaries , and construct the transfer matrices as follows @xmath33 \\ e^{i k_x l_n}(e^{i\theta}+e^{i\phi})/[2e^{-i q_x l_n}\cos \theta ] \\
\end{array } \right ) t_{2n}\end{aligned}\ ] ] where @xmath34 and @xmath35 are the reflection and transmission coefficients of the system that consists of @xmath36 barriers , and @xmath37 is the transfer matrix given by @xmath38 where also @xmath39 is a transfer matrix from site @xmath23 to @xmath40 and @xmath41 are given by @xmath42/2\cos(\varphi_{(i-1)})\\ \nonumber m_{12}&=&e^{-i k_i l_{(i-1)}}e^{-i k_{(i-1)}l_{(i-1 ) } } [ e^{-i\varphi_{(i-1)}}+ e^{-i \varphi_i}]/2\cos(\varphi_{(i-1)})\\ \nonumber m_{21}&=&e^{i k_{i}l_{(i-1)}}e^{i k_{(i-1)}l_{(i-1 ) } } [ e^{i\varphi_{(i-1)}}+ e^{i \varphi_i}]/2\cos(\varphi_{(i-1)})\\ \nonumber m_{22}&=&e^{-i k_{i}l_{(i-1)}}e^{i k_{(i-1)}l_{(i-1 ) } } [ e^{i\varphi_{(i-1)}}- e^{-i \varphi_i}]/2\cos(\varphi_{(i-1)})\end{aligned}\ ] ] where @xmath43 } d_b|_j + int[(i-1)/2]d_w$ ] is the length of system at @xmath23-th boundary and moreover , @xmath44 it is evident that @xmath45 , and that it can be calculated from eq .
( [ eq : t ] ) for a given @xmath36 .
when the transmission coefficients are calculated , the conductivity of system is computed by means of the bttiker formula , @xcite taking the integral of @xmath46 over the angle @xmath47 where @xmath48 with @xmath49 being the width of the graphene strip along the @xmath50 direction .
let us first calculate the transmission probability and study the electronic properties of disordered gsls as a function of the strength of disorder introduced in the system .
we consider the width of barriers as a random variable , so that the length of system in the numerical calculations will be @xmath51 . in all the numerical calculations
, we assumed @xmath52 nm , while the wavelength of the incident particle is set by @xmath53 nm , or , equivalently , the energy of the carrier , @xmath54 mev . in all of the calculations we used , @xmath55 mev , unless otherwise specified .
the number of realization of the random configurations is about 500 .
figure 2 shows the transmission probability , @xmath0 , of the incident electrons hitting a gsls , as a function of the angle @xmath27 for several values of the disorder strength , @xmath24 .
the number of the barriers in the figure is , @xmath56 , with @xmath57 nm .
it is clear that , the transmission decreases by increasing the disorder for all the angles , apart from the strictly normal incidence case , @xmath58 .
this is physically understandable due to the klein tunneling process in graphene , where the backscattering process is suppressed .
moreover , the massless carriers with incident angle close to normal incidence can survive in the presence of disorder , while the width of the angles around the normal incidence decreases with increasing strength of the disorder as well . in order to understand the finite - size effect and the effect of the width of the wells ,
the transmissions probability of a massless particle through the system were calculated as a function of the incident angle .
the results are shown in fig . 3 for several sizes at @xmath59 .
3(a ) we set @xmath57 nm . the transmission decreases with increasing system size for all the angels , except again at @xmath58 .
this behavior is in contrast with a clean gsls result , where the number of the peaks increases by increasing the number of the barriers .
@xcite in fig .
3(b ) the width of the wells is set to , @xmath60 nm . two sharp peaks in the transmission are obtained that disappear in fig .
furthermore , in the case with @xmath61 nm there is only one angle , @xmath62 , that the transmission survives in the presence of the disorder , whereas under other angles the transmissions are suppressed .
meanwhile , there is clearly a wide domain around @xmath58 for which the transmission survives , and is larger than the one shown in fig .
the width of the domain decreases with increasing strength of the disorder .
it is worthwhile to note qualitatively that for @xmath63 , when we have the condition that , @xmath64 ( @xmath65 is an integer ) , the transmission has finite values at angles different from @xmath58 .
this is due to the resonance process in a system with @xmath36 barriers .
we also studied how disorder , introduced in the gsls , affects the conductivity of the system .
hence , we also calculated numerically the dc conductivity by using eq .
( [ eq : dc ] ) , with the white - noise structural disorder imposed on the system .
figure 4 shows the dc conductance of the gsls as a function of @xmath14 for various strengths of the disorder , @xmath24 . as shown in the figure , the conductivity of the gsls decreases by increasing the strength of the disorder .
however , the conductivity approaches a finite value , i.e. , the existence of a finite conductivity in finite - size disordered gsls should be expected . in general , the resonance condition is given by a function that yields , @xmath66 .
for instance , for the case @xmath67 , the condition yields , @xmath68 , as a result of which @xmath69 would be an oscillating function of @xmath70 .
note that @xmath70 is determined by @xmath14 .
consequently , this leads to a finite dc conductivity which is an oscillating function of @xmath14 .
the observation of conductance oscillations in extremely narrow graphene hetrostructures has been observed experimentally .
@xcite in fig .
5 the conductivity of the system is plotted as a function of @xmath14 , where the width of the wells is @xmath71 nm with @xmath59 . the conductivity decreases with increasing the size of system .
the inset in the figure shows the conductivity of a clean gsls as a function of @xmath14 for several system sizes .
it indicates that the dc conductance of clean superlattice behaves uniquely for different sizes , but in a disordered gsls it decreases by increasing the size of system , as shown in fig .
5 . at a constant strength of the disorder ,
changing @xmath12 may also change the conductivity , as depicted in fig .
it demonstrates that the conductivity varies periodically with increasing @xmath12 . as a result , in the disordered gsls
, the dc conductance of finite - size systems depends on the structural parameters , especially @xmath12 . to compute all results that have been presented so far ,
we considered a system of finite size . next , we wish to calculate the finite - size scaling of @xmath72 . for this purpose , we calculated the conductivity as a function of the system size .
the results are summarized in fig .
importantly , the conductivity vanishes by a simple power law , except for the case for which , @xmath73 nm . in general , for @xmath74 the conductivity approaches a finite value as @xmath36 becomes large . in order to examine such results better , we also calculated the @xmath72 for a case for which @xmath75 nm . we found that the conductance tends to a constant in the thermodynamic limit .
the numerical data are fitted by using @xmath76 where @xmath77 , @xmath78 , and @xmath79 are constants .
@xmath77 is the asymptotic value of @xmath72 in the thermodynamic limit , @xmath80 . as a result , for the case @xmath81
nm we obtain @xmath82 , @xmath83 , and @xmath84 ; for @xmath57 nm we obtained , @xmath82 , @xmath85 , and @xmath86 , and @xmath87 , @xmath88 and @xmath89 for @xmath61 nm . in all the case the regression was with @xmath90 , indicating very accurate fits .
note that @xmath77 is zero for small @xmath12 , but tends to a nonzero constant for @xmath61 nm .
we studied numerically the dc conductance of a discorded graphene superlattice p - n junctions for various values of the strength of structural disorder imposed on the material .
it was shown that there exists a width around the normal incidence angle for which the transmission becomes finite in the presence of structural white - noise disorder .
that is , the white - noise disorder gives rise the largest number of the peaks in the transmission , suppressed in the thermodynamic limit but quasiparticles which approach almost perpendicularly to the barriers transmit through the material .
we also calculated the conductivity of a finite - size disordered system and showed that the conductivity decreases by increasing the system size but that there are cases for which the conductance approaches a nonzero value .
this result is in contrast with the case of a clean ( ordered ) gsls .
@xcite furthermore , the results of the finite - size scaling computations predict a zero conductance for all the gsls , except for some special @xmath12 values for which @xmath91 , where @xmath65 is an integer , in which case the conductance tends to a nonzero constant in the thermodynamic limit . apparently , such a feature is independent of the value of @xmath92 . consequently , we predict a finite conductivity for a disordered gsls when the wavelength of incident particle is equal to @xmath93 .
these results are in complete contrast with those calculated for disordered semiconductor superlattice which become insulator . @xcite our finding for the dc conductance of the gsls should be important to the design of electronic nano - devices based on graphene superlattices .
it would probably worthwhile to extend the present work to the case in which a correlated noise is used . in this case
one must replace the white - noise with a proper short- or long - range correlated noise .
r. a. would like to thank the international center for theoretical physics , trieste for its hospitality during the period when part of this work was carried out .
a. e and n. a are supported by the ipm grant .
a. k. geim and k. s. novoselov , nature mater .
* 6 * , 183 ( 2007 ) ; a. k. geim and a. h. macdonald , phys . today * 60 * , 35 ( 2007 ) ; a. h. castro neto , f. guinea , n. m. peres , k. s. novoselov , and a. k. geim , rev . mod
* 81 * , 109 ( 2009 ) .
k. s. novoselov , e. mccann , s. v. morosov , v. falko , m. i. katsnelson , u. zeitler , d. jian , f. scheden , and a. k. geim , nature * 438 * , 197 ( 2005 ) ; y. zhang , j. w. tan , h. l. stormer , and p. kim , nature * 438 * , 201 ( 2005 ) ; z. jiang , e. a. henriksen , l. c. tung , y .- j .
wang , m. e. schwartz , m. y. han , p. kim , and h. l. stormer , phys .
lett . * 98 * , 197403 ( 2007 ) ; a. bostwick , t. ohta , t. seyller , k. horn , and e. rotenberg , nature phys .
* 3 * , 36 ( 2007 ) ; new j. phys .
* 9 * , 385 ( 2007 ) ; m. mucha - kruczyski , o. tsyplyatyev , a. grishin , e. mccann , v. i. falko , a. bostwick , and e. rotenberg , phys .
b * 77 * , 195403 ( 2008 ) ; s. y. zhou , g .- h .
gweon , a. v. fedorov , p. n. first , w. a. de heer , d .- h .
lee , f. guinea , a. h. castro neto , and a. lanzara , nature mat .
* 6 * , 770 ( 2007 ) .
m. polini , r. asgari , g. borghi , y. barlas , t. pereg - barnea and a. h. macdonald , phys .
b * 77 * , 081411(r ) ( 2008 ) ; m. polini , r. asgari , y. barlas , t. pereg - barnea and a. h. macdonald , solid state commun . * 143 * , 58 ( 2007 ) . j. r. williams , l. dicarlo and c. m. marcus , science * 317 * , 638 ( 2007 ) ; b. huard , j. a. sulpizio , n. stander , k. todd , b. yang , and d. goldhaber - gordon , phys
* 98 * , 236803 ( 2007 ) ; b. zyilmaz , p.
jarillo - herrero , d. efetov , d. a. abanin , l. s. levitov and p. kim , phys .
* 99 * , 166804 ( 2007 ) .
s. marchini , s. gunther , j. wintterlin , phys .
rev . b*76 * , 075429 ( 2007 ) ; a. l. vzquez de parga , f. calleja , b. borca , m. c. g. passeggi , jr . , j. j. hinarejos , f. guinea , and r. miranda , phys .
lett . * 100 * , 056807 ( 2008 ) ; i. pletikosi , m. kralj , p. pervan , r. brako , j. coraux , a. t. diaye , c. busse , and t. michely , arxiv:0807.2770 ; y. pan , n. jiang , j. t. sun , d. x. shi , s. x. du , f. liu , and h .- j .
gao , arxiv:0709.2858 .
m. i. katsnelson , eur .
j. b * 51 * , 157 ( 2006 ) ; j. tworzydlo , b. trauzettel , m. titov , a. rycerz , and c. w. beebakker , phys .
lett . * 96 * , 246802 ( 2006 ) ; m. titov , europhys . lett . * 79 * , 17004 ( 2007 ) . of electrons through the system as a function of the incident angle for several disorder strengths .
( a ) : @xmath63 and @xmath94 , and ( b ) : @xmath95 and @xmath96 for @xmath56 and @xmath57nm.,title="fig:",width=117 ] of electrons through the system as a function of the incident angle for several disorder strengths . (
a ) : @xmath63 and @xmath94 , and ( b ) : @xmath95 and @xmath96 for @xmath56 and @xmath57nm.,title="fig:",width=257 ] of the massless carriers through the system as a function of the incident angle , the system size and @xmath59 for @xmath57 nm .
( a ) @xmath97 nm .
( b ) @xmath98 nm.,title="fig:",width=6 ] of the massless carriers through the system as a function of the incident angle , the system size and @xmath59 for @xmath57 nm .
( a ) @xmath97 nm .
( b ) @xmath98 nm.,title="fig:",width=6 ] of the massless carriers through the system as a function of the incident angle , the system size and @xmath59 for @xmath57 nm . ( a ) @xmath97 nm . ( b ) @xmath98 nm.,title="fig:",width=6 ] | we study the conductance of disordered graphene superlattices with short - range structural correlations .
the system consists of electron- and hole - doped graphenes of various thicknesses , which fluctuate randomly around their mean value .
the effect of the randomness on the probability of transmission through the system of various sizes is studied .
we show that in a disordered superlattice the quasiparticle that approaches the barrier interface almost perpendicularly transmits through the system . the conductivity of the finite - size system is computed and shown that the conductance vanishes when the sample size becomes very large , whereas for some specific structures the conductance tends to a nonzero value in the thermodynamics limit .
= 0.55 cm |
one of the motivations of this paper is to discuss some `` mysterious '' configurations of zeros of polynomials , defined by an orthogonality condition with respect to a sum of exponential functions on the plane , that appeared as a results of our numerical experiments .
it turned out that in this apparently simple situation the orthogonal polynomials may exhibit a behavior which existing theoretical models do not explain , or the explanation is not straightforward . in order to make our arguments
self - contained , we present a brief outline of the fundamental concepts and known results and discuss their possible generalizations . the so - called complex or non - hermitian
orthogonal polynomials with analytic weights appear in approximation theory as denominators of rational approximants to analytic functions @xcite and in the study of continued fractions .
recently , non - hermitian orthogonality found applications in several new areas , for instance in the description of the rational solutions to painlev equations @xcite , in theoretical physics @xcite and in numerical analysis @xcite .
observe that due to analyticity , there is a freedom in the choice of the integration contour for the non - hermitian orthogonal polynomials , which means that the location of their zeros is a priori not clear .
the problem of their limit zero distribution is one of the central aspects studied in the theory of orthogonal polynomials , especially in the last few decades .
several important general results in this direction have been obtained , and describing them is one of the goals of the first parts of this paper . however , the general theory is far from being complete , and many natural questions remain unanswered or have only a partial explanation , as some of the examples presented in the second part of this work will illustrate .
we will deal with one of the simplest situation that is still posing many open questions .
complex non - hermitian orthogonal polynomials are denominators of the diagonal pad approximants to functions with branch points and thus play a key role in the study of the asymptotic behavior of these approximants , in particular , in their convergence .
since the mid - twentieth century convergence problems for pad approximants have been attracting wide interest , and consequently , complex orthogonal polynomials have become one of the central topics in analysis and approximation theory .
there is a natural historical parallel of this situation with the one occurred in the middle of the ninetieth century , when pad approximants ( studied then as continued fractions ) for markov- and stieltjes - type functions led to the introduction of general orthogonal polynomials on the real line .
the original fundamental theorems by p. chebyshev , a. markov and t. stieltjes on the subject gave birth to the theory of general orthogonal polynomials . in 1986 stahl @xcite proved a fundamental theorem explaining the geometry of configurations of zeros of non - hermitian orthogonal polynomials and presented an analytic description of the curves `` drawn '' by the strings of zeros .
those curves are important particular cases of what we now call @xmath0-curves .
they may be defined by the symmetry property of their green functions or as trajectories of some quadratic differential .
the fact that the denominators of the diagonal pad approximants to an analytic function at infinity satisfy non - hermitian orthogonality relations is straightforward and was definitely known in the nineteenth century .
just nobody believed that such an orthogonality could be used to study the properties pad denominators .
stahl s theorem showed that complex orthogonality relations may be effectively used for these purposes , at least for functions with a `` small '' set of singular points , some of them being branch points .
this , without any doubt , was a beginning of a new theory of orthogonal polynomials . before the work of stahl , asymptotics of these polynomials was studied for some subclasses of functions and by appealing to their additional properties .
for instance , several important results were obtained by gonchar @xcite and collaborators in 1970s and the beginning of the 1980s .
the geometry of their zero distribution was conjectured ( and in partially proved , e.g. , for hyperelliptic functions ) by j. nuttall and collaborators @xcite .
later , in @xcite , the case when the logarithmic derivative of the approximated function is rational ( the so - called semiclassical or laguerre class ) was analyzed .
the associated orthogonal polynomials , known as the semiclassical or generalized jacobi polynomials , satisfy a second - order differential equation , and the classical liouville
green ( a.k.a wkb ) method may be used to study their strong asymptotics , as it was done by nuttall , see also a recent paper @xcite .
stahl s ideas were considerably extended by gonchar and rakhmanov @xcite to cover the case when the orthogonality weight depends on the degree of the polynomial , which requires the inclusion of a non - trivial external field ( or background potential ) in the picture .
the curves , describing the location of the strings of zeros of the orthogonal polynomials , feature a symmetry property ( the @xmath0-property , so we call them the @xmath0-curves ) , and their geometry is much more involved . the resulting gonchar rakhmanov stahl ( or grs ) theory , founded by @xcite , allows to formulate statements about the asymptotics of the zeros of complex orthogonal polynomials conditional to the existence of the @xmath0-curves , which is a non - trivial problem from the geometric function theory
. further contributions in this direction , worth mentioning here , are @xcite .
the notion of the @xmath0-property can be interpreted also in the light of the deift - zhou s nonlinear steepest descent method for the riemann
hilbert problems @xcite .
one of the key steps in the asymptotic analysis is the deformation of the contours , the so - called lens opening , along the level sets of certain functions .
it is precisely the @xmath0-property of these sets which guarantees that the contribution on all non - relevant contours becomes asymptotically negligible .
an important further development was a systematic investigation of the critical measures , presented in @xcite .
critical measures are a wider class that encompasses the equilibrium measures on @xmath0-curves , see sect . [ sec : grs ] for the precise definition and further details .
one of the contributions of @xcite was the description of their supports in terms of trajectories of certain quadratic differentials ( this description for the equilibrium measures with the @xmath0-property is originally due to stahl @xcite ) . in this way
, the problem of existence of the appropriate @xmath0-curves is reduced to the question about the global structure of such trajectories .
let us finish by describing the content of this paper .
section [ sec : examples ] is a showcase of some zero configurations of polynomials of complex orthogonality , appearing in different settings .
the presentation is mostly informal , it relies on some numerical experiments , and its goal is mainly to illustrate the situation and eventually to arouse the reader s curiosity .
[ sec : logpotential ] contains a brief overview of some basic definitions from the logarithmic potential theory , necessary for the subsequent discussion , as well as some simple applications of these notions to polynomials .
this section is essentially introductory , and a knowledgeable reader may skip it safely . in sect .
[ sec : grs ] we present the known basic theorem on asymptotics of complex orthogonal polynomials .
we simplify settings as much as possible without losing essential content .
the definitions and results contained here constitute the core of what we call the _
grs theory_. altogether , sects . [ sec : examples][sec : grs ] are expository . finally , in sections [ subsec6.7 ] and [ sec:6 ] we present some recent or totally new results . for instance
, section [ subsec6.7 ] is about the so - called vector critical measures , which find applications in the analysis of the hermite
pad approximants of the second kind for a couple of power series at infinity of a special form , as well as in the study of the problems tackled in sect .
[ sec:6 ] .
this last section of the paper deals with the orthogonality with respect to _ a sum _ of two ( or more ) analytic weights . in order to build some intuition
, we present another set of curious numerical results in sect .
[ sec : numerical experiments ] , and the title of this paper ( partially borrowed from philip k. dick ) is motivated by the amazing variety and beauty of possible configurations . as the analysis of these experiments shows even for the simplest model , corresponding to the sum of two exponential weights , in some domain in the parameter space of the problem the standard grs theory still explains the observed behavior , while in other domains it needs to be modified or adapted , which leads to some new equilibrium problems
. finally , there are regions in the parameter space where it is not yet clear how to generalize the grs theory to explain our numerical results , and we can not go beyond an empirical discussion .
we start by presenting some motivating examples that should illustrate the choice of the title of this work .
pad approximants @xcite are the _ locally best rational approximants of a power series _ ; in a broader sense , they are constructive rational approximants _ with free poles_. let @xmath1 denote the set of algebraic polynomials with complex coefficients and degree @xmath2 , and let @xmath3 be a ( formal ) power series .
for any arbitrary non - negative integer @xmath4 there always exist polynomials @xmath5 and @xmath6 , @xmath7 , satisfying the condition @xmath8 this equation is again formal and means that @xmath9 ( called the _ remainder _ ) is a power series in descending powers of @xmath10 , starting at least at @xmath11 . in order to find polynomials @xmath12 and @xmath13 we first use condition to determine the coefficients of @xmath13 , after which @xmath12 is just the truncation of @xmath14 at the terms of nonnegative degree .
it is easy to see that condition does not determine the pair @xmath15 uniquely .
nevertheless , the corresponding rational function @xmath16 is unique , and it is known as the _ ( diagonal ) pad approximant to @xmath17 at @xmath18 of degree @xmath4_. hence , the denominator @xmath13 is the central object in the construction of diagonal pad approximants , and its poles constitute the main obstruction to convergence of @xmath19 in a domain of the complex plane @xmath20 . with this definition we can associate a formal orthogonality verified by the denominators @xmath13 ( see , e.g. , a recent survey @xcite and the references therein ) .
however , the most interesting theory is developed when @xmath17 is an analytic germ at infinity .
indeed , if converges for @xmath21 , then choosing a jordan closed curve @xmath22 in @xmath23 and using the cauchy theorem we conclude that @xmath24 this condition is an example of a _ non - hermitian orthogonality _
satisfied by the denominators @xmath13 . in particular interesting
is the case when @xmath17 corresponds to an algebraic ( multivalued ) function , being approximated by intrinsically single - valued rational functions @xmath19 . as an illustration we plot in figure [ fig1pade ] the poles of @xmath25 for the analytic germs at infinity of two functions , @xmath26 both normalized by @xmath27 .
these functions belong to the so - called laguerre class ( or are also known as `` semiclassical '' ) : their logarithmic derivatives are rational functions .
cc fig1a & fig1b a quick examination of the pictures puts forward two phenomena : 1 .
generally , poles of @xmath19 distribute on @xmath20 in a rather regular way .
our eye can not avoid `` drawing '' curves along which the zeros align almost perfectly .
2 . there are some exceptions to this beautiful order : observe a clear outlier on figure [ fig1pade ] , right .
these `` outliers '' are known as the _ spurious poles _ of the pad approximants .
this is the `` most classical '' family of polynomials , which includes the chebyshev polynomials as a particular case .
they can be defined explicitly ( see @xcite ) , @xmath28 or , equivalently , by the well - known rodrigues formula @xmath29 .\ ] ] incidentally , jacobi polynomials could have been considered in the previous section : for @xmath30 they are also denominators of the diagonal pad approximants ( at infinity ) to the function @xmath31 in fact , denominators of the diagonal pad approximants to semiclassical functions as in are known as _ generalized jacobi polynomials _ , see @xcite . clearly , polynomials @xmath32 are entire functions of the complex parameters @xmath33 .
when @xmath30 they are orthogonal on @xmath34 $ ] with respect to the positive weight @xmath35 , @xmath36 and in consequence their zeros are all real , simple , and belong to @xmath37 . what happens if at least one of the parameters @xmath38 is `` non - classical '' ? in figure [ fig : onlyzeros ] we depicted the zeros of @xmath32 for @xmath39 , @xmath40 and @xmath41 .
we can appreciate again the same feature : a very regular distribution of the zeros along certain imaginary lines on the plane .
fig2(29,31)@xmath42 ( 68,33)@xmath43 these are a natural generalization of the jacobi polynomials . given a set of pairwise distinct points fixed on the complex plane @xmath20 , @xmath44 ( @xmath45 ) , and two polynomials , @xmath46 we are interested in the polynomial solutions of the _ generalized lam differential _ equation ( in algebraic form ) , @xmath47 where @xmath48 ; if @xmath49 , then @xmath50 is monic .
an alternative perspective on the same problem can be stated in terms of the second order differential operator @xmath51(z ) : = a(z)\ , y''(z)+ b(z)\ , y'(z),\ ] ] and the associated generalized spectral problem ( or multiparameter eigenvalue problem , see @xcite ) , @xmath52(z)=n(n+\alpha -1 ) v_n(z)\ , y(z)\ , , \quad n \in { \mathbb{n}},\ ] ] where @xmath53 is the `` spectral polynomial '' .
if we take @xmath54 in , we are back in the case of jacobi polynomials ( hypergeometric differential equation ) . for @xmath55
we get the heun s equation , which still attracts interest and poses open questions ( see @xcite ) .
moreover , denominators of pad approximants for semiclassical functions are also heine stieltjes polynomials , see e.g. @xcite .
heine @xcite proved that for every @xmath56 there exist at most @xmath57 different polynomials @xmath58 such that ( or ) admits a polynomial solution @xmath59 .
these particular @xmath58 are called _ van vleck polynomials _ , and the corresponding polynomial solutions @xmath60 are known as _ heine - stieltjes _ ( or simply stieltjes ) polynomials ; see @xcite for further details ( fig .
[ fig : zeroshs ] ) .
fig3 stieltjes discovered an electrostatic interpretation of zeros of the polynomials discussed in @xcite , which attracted general attention to the problem .
he studied the problem in a particular setting , assuming that @xmath61 and that all residues @xmath62 in @xmath63 are strictly positive ( which is equivalent to the assumption that the zeros of @xmath64 alternate with those of @xmath65 and that the leading coefficient of @xmath65 is positive ) .
he proved in @xcite ( see also ( * ? ? ?
* theorem 6.8 ) ) that in this case for each @xmath56 there are _ exactly _ @xmath66 different van vleck polynomials of degree @xmath67 and the same number of corresponding heine - stieltjes polynomials @xmath68 of degree @xmath4 , given by all possible ways how the @xmath4 zeros of @xmath68 can be distributed in the @xmath69 open intervals defined by @xmath70 . obviously , this models applies also in the case @xmath54 , i.e. to the zeros of a jacobi polynomial . for a more detailed discussion
see @xcite , although we describe the electrostatic model of stieltjes for @xmath54 in the next section .
let us return to the pad approximants . the situation gets more involved if we consider now a _
pair _ of power series at infinity , say @xmath71 fix a non - negative integer @xmath4 , and seek _ three _ polynomials , @xmath72 , @xmath73 and @xmath74 , with all @xmath75 and not all @xmath76 , such that @xmath77 @xmath78 are the _ type i hermite pad ( hp ) polynomials _ , corresponding to the pair @xmath79 ( or more precisely , to the vector @xmath80 ) , and function @xmath9 defined in is again the remainder of the hermite
pad approximation to @xmath81 .
hermite used in 1858 a construction slightly more general than , involving @xmath82 , in order to prove that the number @xmath83 is transcendental .
hp polynomials play an important role in analysis and have significant applications in approximation theory , number theory , random matrices , mathematical physics , and other fields . for details and further references see @xcite .
fig4 hp polynomials is another classical construction closely related to orthogonal polynomials , but essentially more complicated .
an asymptotic theory for such polynomials is not available yet , even though there is a number of separate results . as a single illustration of the sophisticated beauty and complexity of this situation , we plot in figure [ fig1hermitepade ] the zeros of @xmath84 for the analytic germs at infinity of two functions , @xmath85 both normalized by @xmath86 . as in , functions @xmath87 and @xmath88 are semiclassical , and it is known @xcite that polynomials
@xmath89 satisfy a differential equation .
observe the interesting and non - trivial configuration of the curves where their zeros lie . at any rate ,
let us insist in a general conclusion we can extract from all previous examples : zeros of analyzed polynomials tend to distribute along certain curves on @xmath20 in a very regular way that clearly needs an explanation .
how can we count the zeros of a polynomial ?
a trivial observation is that if @xmath90 is a monic polynomial of degree @xmath4 , then its logarithmic derivative @xmath91 can be written as @xmath92 where @xmath93 is the _ normalized zero - counting measure _ for @xmath69 . observe that we count each zero in accordance with its multiplicity .
the integral in the right - hand side of is related to the so - called _ cauchy _ or _ stieltjes transform _ of this measure . in general , given a finite borel measure @xmath94 on @xmath20 , its cauchy transform ( in the sense of the principal value ) is @xmath95 hence , our first identity is @xmath96 which is valid as long as @xmath97 , or equivalently , for @xmath98 . the second
, apparently trivial observation is that @xmath99 for a finite borel measure @xmath94 on @xmath20 we define its logarithmic potential @xmath100 hence , latexmath:[\[\label{identity2 } @xmath98 .
the apparently innocent identities establish the connection between the analytic properties of polynomials and two very important areas of analysis : the harmonic analysis and singular integrals , and the logarithmic potential theory .
let us start with a classical example : the zero distribution of the jacobi polynomials , defined in .
an already mentioned fact is that the jacobi polynomial @xmath102 is solution of a second - order differential equation ( hypergeometric equation ) @xmath103 where @xmath104 , see ( * ? ? ?
* theorem 4.2.2 ) . for @xmath30
all zeros of @xmath105 are real , simple , and belong to @xmath37 . with @xmath30 and @xmath106
, consider the normalized zero - counting measures @xmath107 .
standard arguments using weak compactness of the sequence @xmath108 show that there exist @xmath109 and a unit measure @xmath110 on @xmath34 $ ] such that @xmath111 here we denote by @xmath112 the weak- * convergence of measures .
an expression for the cauchy transform @xmath113 of @xmath110 can be obtained in an elementary way directly from ( [ difjac ] ) .
the derivation of the continued fraction for @xmath114 from the differential equation appears in the perron s monograph @xcite , although the original ideas are contained already in the work of euler . for more recent applications , see @xcite .
the differential equation can be rewritten in terms of the function @xmath115 well defined at least for @xmath116 $ ] .
we get @xmath117 since zeros of @xmath118 and @xmath119 interlace on @xmath37 , functions @xmath120 are analytic and uniformly bounded in @xmath121 $ ] , and by our assumption , @xmath122 uniformly on compact subsets of ( a.k.a . locally uniformly in ) @xmath121 $ ] .
thus , taking limits in ( [ diftemp ] ) , we obtain that @xmath114 is an algebraic function satisfying a very simple equation : @xmath123 in other words , @xmath124,\ ] ] where we take the branch of the square root satisfying @xmath125 for @xmath126 . in possession of the additional information that @xmath127 $ ] we can use the stieltjes perron ( or sokhotsky plemelj ) inversion formula to recover the measure @xmath110 : we conclude that @xmath110 is absolutely continuous on @xmath34 $ ] , @xmath128 $ ] , and @xmath129.\ ] ] due to the uniqueness of this expression , we conclude that the limit in holds for the whole sequence @xmath130 .
limit with @xmath110 given by holds actually for families of orthogonal polynomials with respect to a wide class of measures @xmath94 on @xmath34 $ ] .
the explanation lies in the properties of the measure @xmath110 : it is the _ equilibrium measure _ of the interval @xmath34 $ ] . in order to define it properly
, we need to introduce the concept of the _ logarithmic energy _ of a borel measure @xmath131 : @xmath132 moreover , given a real - valued function @xmath133 ( the _ external field _ ) , we consider also the _ weighted energy _ @xmath134 the electrostatic model of stieltjes for the zeros of the jacobi polynomials @xmath135 says precisely that the normalized zero - counting measure @xmath136 minimizes @xmath137 , with @xmath138 among all discrete measures of the form @xmath139 supported on @xmath34 $ ] .
notice that @xmath140 vanishes asymptotically as @xmath106 for @xmath141 , so that it is not surprising that the weak- * limit @xmath110 of @xmath108 , given by on @xmath34 $ ] , minimizes the logarithmic energy among all probability borel measures living on that interval : @xmath142 $ , and } \int d\mu=1 \right\}.\ ] ] in the terminology of potential theory , @xmath94 is the _ equilibrium measure _ of the interval @xmath34 $ ] , the value @xmath143 is its _ robin constant _ , and @xmath144)=\exp(-e(\lambda))=\frac{1}{2}\ ] ] is its _
logarithmic capacity_. as it was mentioned , this asymptotic zero behavior is in a sense universal : it corresponds not only to @xmath135 , but to any sequence of orthogonal polynomials on @xmath34 $ ] with respect to a measure @xmath145 a.e .
( see , e.g. , @xcite ) , or even with respect to complex measures with argument of bounded variation and polynomial decay at each point of its support @xcite .
we no longer have a differential equation , but we do have the @xmath146 minimal property of the orthogonal polynomial , which at the end suffices .
notice that the arguments above extend easily to the case of parameters @xmath38 dependent on @xmath4 .
for instance , when the limits @xmath147 exist , the * -limit of the zero - counting measure of @xmath148 minimizes the weighted energy with a non - trivial external field @xmath149 ( see @xcite for a general treatment of this case ) .
in this situation the sequence @xmath148 satisfies _ varying orthogonality _ conditions , when the weight in depends on the degree of the polynomial .
as it was shown by gonchar and rakhmanov @xcite , under mild assumptions same conclusions will be valid for general sequences of polynomials satisfying varying orthogonality conditions .
but let us go back to the zeros of an individual jacobi polynomial @xmath135 . under the assumption of @xmath30 ,
why do they belong to the real line ?
a standard explanation invokes orthogonality , but why are we integrating along the real line ?
the integrand in is analytic , so any deformation of the integration path joining @xmath150 and @xmath151 leaves the integral unchanged .
why do the zeros still go to @xmath152 ? we can modify stieltjes electrostatic model to make it `` @xmath152-free '' ( see @xcite ) : on a continuum @xmath153 joining @xmath150 and @xmath151 find a discrete measures of the form @xmath139 , supported on @xmath153 , minimizing @xmath137 , with @xmath154 as in ; the minimizer is not necessarily unique .
denote this minimal value by @xmath155 .
now maximize @xmath155 among all possible continua @xmath153 joining @xmath150 and @xmath151 .
the resulting value is @xmath156 , and the max
min configuration is on @xmath37 , given by the zeros of @xmath135 .
min ansatz remains valid in the limit @xmath106 : among all measures supported on a continuum joining @xmath150 and @xmath151 , @xmath110 in maximizes the minimal energy . or equivalently , @xmath34 $ ] _ is the set of minimal logarithmic capacity _ among all continua joining @xmath150 and @xmath151 . if we recall that the jacobi polynomials are denominators of the diagonal pad approximants ( at infinity ) to the function , using markov s theorem ( see @xcite ) we can formulate our conclusion as follows : the diagonal pad approximants to this @xmath17 converge ( locally uniformly ) in @xmath157 , where @xmath153 is the continuum of minimal capacity joining @xmath150 and @xmath151
it is easy to see that @xmath17 is a multivalued analytic function with branch points at @xmath158 .
it turns out that this fact is much more general , and is one of the outcomes of the gonchar
stahl ( or grs ) theory .
let us denote by @xmath159 the class of functions holomorphic ( i.e. , analytic and single - valued ) in a domain @xmath160 , @xmath161 and let @xmath162 be defined by the minimal capacity property @xmath163 stahl @xcite proved that the sequence @xmath164 converges to @xmath17 in capacity in the complement to @xmath0 , @xmath165 under the assumption that the set of singularities of @xmath17 has capacity @xmath166 .
the convergence in capacity ( instead of uniform convergence ) is the strongest possible assertion for an arbitrary function @xmath17 due to the presence of the so - called spurious poles of the pad approximants that can be everywhere dense , even for an entire @xmath17 , see @xcite , as well as figure [ fig1pade ] .
more precisely , stahl established the existence of a unique @xmath167 of minimal capacity , comprised of a union of analytic arcs , such that the jump @xmath168 across each arc is @xmath169 , as well as the fact that for the denominator @xmath13 of pad approximants @xmath19 we have @xmath170 , where @xmath171 is equilibrium measure for @xmath0 . here and in what follows
we denote by @xmath172 ( resp . ,
@xmath173 ) the left ( resp . , right )
boundary values of a function @xmath17 on an oriented curve .
the original work of stahl contained not only the proof of existence , but a very useful characterization of the extremal set @xmath0 : on each arc of this set @xmath174 where @xmath175 are the normal vectors to @xmath0 pointing in the opposite directions .
this relation is known as the _ @xmath0-property _ of the compact @xmath0 .
notice that stahl s assertion is not conditional , and the existence of such a compact set of minimal capacity is guaranteed . in the case of a finite set of singularities ,
the simplest instance of such a statement is the content of the so - called _
chebotarev s problem _ from the geometric function theory about existence and characterization of a continuum of minimal capacity containing a given finite set .
it was solved independently by grtzsch and lavrentiev in the 1930s .
a particular case of stahl s results , related to chebotarev s problem , states that given a finite set @xmath176 of @xmath177 distinct points in @xmath20 there exist a unique set @xmath0 @xmath178 where @xmath179 is the class of continua @xmath180 with @xmath181 .
the complex green function @xmath182 for @xmath183 has the form @xmath184 where @xmath185 and @xmath50 is a polynomial uniquely defined by @xmath70 . in particular , we have @xmath186 and is an immediate consequence of these expressions .
another consequence is that @xmath0 is a union of arcs of critical trajectories of the quadratic differential @xmath187 .
this is also the zero level of the ( real ) green function @xmath188 of the two - sheeted riemann surface for @xmath189 . in order to study the limit zero distribution of pad denominators
@xmath13 stahl @xcite created an original potential theoretic method based directly on the non - hermitian orthogonality relations @xmath190 satisfied by these polynomials ; incidentally , he also showed for the first time how to deal with a non - hermitian orthogonality .
the method was further developed by gonchar and rakhmanov in @xcite for the case of varying orthogonality ( see also @xcite ) . the underlying potential theoretic model in this case
must be modified by including a non trivial external field @xmath191 .
if the set @xmath0 on the plane is comprised of a finite number of piece - wise analytic arcs , we say that it exhibits the _
@xmath0-property in the external field @xmath191 _ if @xmath192 where @xmath193 is now the minimizer of the weighted energy among all probability measures supported on @xmath0 , and @xmath194 .
in other words , @xmath110 is the equilibrium measure of @xmath0 in the external field @xmath191 , and can be characterized by the following variational ( equilibrium ) conditions : there exists a constant @xmath195 ( the _ equilibrium constant _ ) such that @xmath196 equation uniquely defines both the probability measure @xmath197 on @xmath0 and the constant @xmath198 .
the pair of conditions has a standard electrostatic interpretation , which turns useful for understanding the structure of the @xmath0-configurations .
indeed , it follows from that distribution of a positive charge presented by @xmath110 is in equilibrium on the fixed conductor s , while the @xmath0-property of compact @xmath0 in means that forces acting on the element of charge at @xmath199 from both sides of @xmath0 are equal .
so , the distribution @xmath110 of an @xmath0-curve will remain in equilibrium if we remove the condition ( `` scaffolding '' ) that the charge belongs to @xmath0 and make the whole plane a conductor ( except for a few exceptional insulating points , such as the endpoints of some of arcs in the support of @xmath110 ) . in other words , @xmath110 is a distribution of charges which is in an ( unstable ) equilibrium in a conducting domain .
let @xmath160 be a domain in @xmath20 , @xmath153 a compact subset of @xmath160 of positive capacity , @xmath200 , and let the sequence @xmath201 .
assume that polynomials @xmath202 are defined by the non - hermitian orthogonality relations @xmath203 where @xmath204 the integration goes along the boundary of @xmath205 ( if such integral exists , otherwise integration goes over an equivalent cycle in @xmath205 ) . a slightly simplified version of one of the main results of gonchar and rakhmanov @xcite is the following : [ grsthm ] assume that @xmath206 converge locally uniformly in @xmath160 ( as @xmath106 ) to a function @xmath207 .
if @xmath153 has the @xmath0-property in @xmath208 and if the complement to the support of the equilibrium measure @xmath193 is connected , then @xmath209 .
theorem [ grsthm ] was proved in @xcite , where it was called `` generalized stahl s theorem '' .
observe that unlike the original stahl s theorem , its statement is conditional : _ if _ we are able to find a compact set with the @xmath0-property ( in a harmonic external field ) and connected complement , _ then _ the weak- * convergence is assured . under general assumptions on the class of integration paths @xmath210 and in the presence of a non trivial external field , neither existence nor uniqueness of a set with the @xmath0-property are guaranteed .
a general method of solving the existence problem is based on the maximization of the equilibrium energy , which is inspired by the minimum capacity characterization ( in the absence of an external field ) , see .
more exactly , consider the problem of finding @xmath211 with the property @xmath212 where @xmath213 is the set of all probability borel measures on @xmath214 .
if a solution @xmath0 of this extremal problem exists then under `` normal circumstances '' @xmath0 is an @xmath0-curve in the external field @xmath191 , see @xcite . _ critical measures _ are borel measures on the plane with a vanishing local variation of their energy in the class of all smooth local variations with a prescribed set of fixed points .
they are , therefore equilibrium distributions in a conducting plane with a number of insulating points of the external field .
the zeros of the heine stieltjes polynomials ( see section [ sec : hspolyns ] ) are ( discrete ) critical measures .
this observation lead in @xcite to a theorem on asymptotics of these polynomials in terms of continuous critical measures with a finite number of insulating points of the external field .
it turns out that the equilibrium measures of compact sets with the @xmath0-property are critical ; the reciprocal statement is that the critical measures may be interpreted as the equilibrium measures of @xmath0-curves in piece - wise constant external fields .
both notions , however , are defined in a somewhat different geometric settings , and it is in many ways convenient to distinguish and use both concepts in the study of many particular situations .
the idea of studying critical ( stationary ) measures has its origins in @xcite .
later it was used in @xcite in combination with the min
max ansatz , and systematically in @xcite , see also @xcite .
the formal definition is as follows : let @xmath215 be a domain , @xmath216 be a finite set and @xmath191 a harmonic function in @xmath217 .
let @xmath218 and @xmath219 , satisfying @xmath220 .
then for any ( signed ) borel measure @xmath94 the mapping ( `` @xmath221-variation '' ) @xmath222 defines the pull - back measure @xmath223 , as well as the associated variation of the weighted energy , @xmath224 where @xmath225 was introduced in .
we say that @xmath94 is _ @xmath226-critical _
if for any @xmath221-variation , such that the limit above exists , we have @xmath227 when @xmath228 , we write @xmath221-critical instead of @xmath229-critical measure . the relation between the critical measures and the @xmath0-property is very tight : every equilibrium measures with an @xmath0-property is critical , while the potential of any critical measure satisfies .
however , in some occasion it is more convenient to analyze the larger set of critical measures .
as it was proved in @xcite , for any @xmath226-critical measure @xmath94 we have @xmath230 this formula implies the following description of @xmath231 : it consists of a finite number of critical or closed trajectories of the quadratic differential @xmath232 on @xmath20 ( moreover , all trajectories of this quadratic differential are either closed or critical , see ( * ? ? ?
* theorem 5.1 ) or @xcite ) . together with this
yields the representation @xmath233 finally , the @xmath0-property on @xmath153 and the formula for the density @xmath234 on any open arc of @xmath153 follow directly from . in this way
, function @xmath235 becomes the main parameter in the construction of a critical measure : if we know it ( or can guess it correctly ) , then consider the problem solved .
[ examplesaffconf ] let us apply the grs theory to the following simple example : we want to study zeros asymptotics of the polynomials @xmath236 defined by the orthogonality relation @xmath237 where the varying ( depending on @xmath4 ) weight function has the form @xmath238 and the integration goes along a jordan arc @xmath214 connecting @xmath239 and @xmath240 .
since @xmath241 uniformly on every compact subset of @xmath20 , the application of the grs theory is reduced to finding a jordan arc @xmath0 , connecting @xmath242 and @xmath243 , such that the equilibrium measure @xmath193 has the @xmath0-property in the external field @xmath191 . for @xmath244 , such an @xmath0 is just the vertical straight segment connecting both endpoints . using and the results of @xcite
it is easy to show that for small values of @xmath245 ( roughly speaking , for @xmath246 ) , @xmath247 valid a.e .
( with respect to the plane lebesgue measure ) on @xmath20 . in this case
, @xmath248 is connected , and is the critical trajectory of the quadratic differential @xmath249 connecting @xmath250 . as an illustration of this case we depicted the zeros of @xmath251 for @xmath252 , see figure [ figexampleexp ] , left . ] .
cc fig5a & fig5b at a critical value of @xmath253 the double zero @xmath254 hits the critical trajectory and splits ( for @xmath255 ) into two simple zeros , that can be computed , so that now @xmath256 in this situation @xmath257 is made of two real - symmetric open arcs , each connecting @xmath258 with @xmath259 lying in the same half - plane , delimited by @xmath152 ; they are the critical trajectory of the quadratic differential @xmath260 on @xmath20 . for illustration , see the zeros of @xmath251 for @xmath261 , see figure [ figexampleexp ] , right .
formula and the fact that @xmath153 is comprised of arcs of trajectories of @xmath262 on @xmath20 show that @xmath263 is a constant on @xmath153 . however , this constant is not necessarily the same on each connected component of @xmath153 : this additional condition singles out the equilibrium measures with the @xmath0-property within the class of critical measures .
equivalently , the equilibrium measure can be identified by the validity of the variational condition .
recall that one of the motivations for the study of critical ( instead of just equilibrium ) measures is the characterization of the zero distributions of heine stieltjes polynomials , see section [ sec : hspolyns ] .
one of the main results in @xcite is the following .
if we have a convergent sequence of van vleck polynomials @xmath264 , then for corresponding heine
stieltjes polynomials @xmath13 we have @xmath265 where @xmath94 is an @xmath221-critical measure .
moreover , any @xmath221-critical measure may be obtained this way . in electrostatic terms
, @xmath266 is a discrete critical measure and the result simply means that a weak limit of a sequence of discrete critical measures is a ( continuous ) critical measure . in the case of the heine
stieltjes polynomials , @xmath235 in is a rational function of the form @xmath267 , with @xmath64 given by , and the polynomial @xmath22 determined by a system of nonlinear equations . in the case of varying jacobi polynomials , @xmath148 , with @xmath268 ,
@xmath269 satisfying @xmath270 we can obtain the explicit expression for @xmath235 using the arguments of section [ sec : logpotential ] . by the grs theory , the problem of the weak- * asymptotics of the zeros of such polynomials boils down to the proof of the existence of a critical trajectory of the corresponding quadratic differential , joining the two zeros of @xmath22 , and of the connectedness of its complement in @xmath20 , see @xcite , as well as figure [ fig : withzeros ] .
if we check the motivating examples from section [ sec : examples ] , we will realize that at this point we only lack tools to explain the asymptotics of the zeros of the hermite
pad polynomials ( section [ sec : hpapprox ] ) . for this
, we need to extend the notion of critical measures ( and in particular , of equilibrium measures on sets with an @xmath0-property ) to a vector case .
assume we are given a vector of @xmath271 nonnegative measures @xmath272 , compactly supported on the plane , a symmetric and positive - semidefinite interaction matrix @xmath273 , and a vector of real - valued harmonic external fields @xmath274 , @xmath275 , @xmath276 .
we consider the total ( vector ) energy functional of the form @xcite @xmath277 ( compare with ) , where @xmath278 is the mutual logarithmic energy of two borel measures @xmath94 and @xmath279 .
typical cases of the matrix @xmath280 for @xmath55 are @xmath281 corresponding to the so - called angelesco and nikishin systems , respectively , see @xcite . as in the scalar situation , for @xmath218 and @xmath282 , denote by @xmath283 the push - forward measure of @xmath284 induced by the variation of the plane @xmath285 , @xmath286 .
we say that @xmath287 is a
_ critical vector measure _ ( or a saddle point of the energy @xmath288 ) if @xmath289 for every function @xmath282 .
usually , critical vector measures a sought within a class specified by their possible support and by some constraints on the size of each component .
the vector _ equilibrium _ problems deal with the minimizers of the energy functional over such a family of measures @xmath287 .
for instance , we can be given @xmath69 families of analytic curves @xmath290 , so that @xmath291 , @xmath292 , and additionally some constraints on the size of each component @xmath293 of @xmath294 . in a classical setting
this means that we fix the values @xmath295 , such that @xmath296 , and impose the condition @xmath297 see e.g. @xcite .
more recently new type of constraints have found applications in this type of problems , when the conditions are imposed on a linear combination of @xmath298 .
for instance , @xcite considers the case of @xmath299 , with the interaction matrix @xmath300 polynomial external fields , and conditions @xmath301 $ ] is a parameter ; see also @xcite and @xcite for @xmath302 and its applications to hermite - pad approximants . from the electrostatic point of view
it means that we no longer fix the total masses of each component @xmath293 , but charges ( of either sign ) can `` flow '' from one component to another .
there is a natural generalization of the scalar @xmath0-property to the vector setting .
continuing with the results of the recent work @xcite , it was proved that under some additional natural assumptions on the critical vector measure @xmath294 , if @xmath303 is an open analytic arc in @xmath304 , for some @xmath305 , then there exists a constant @xmath306 for which both the variational equation ( _ equilibrium condition _ ) @xmath307 and the @xmath0-property @xmath308 hold true a.e . on @xmath309 , where again @xmath310 are the unit normal vectors to @xmath303 , pointing in the opposite directions
. clearly , the problem of existence of compact sets with the @xmath0-property for the vector problem are even more difficult than in the scalar case .
the possibility to reduce this question to the global structure of critical trajectories of some quadratic differentials is the most valuable tool we have nowadays .
this is given by higher - dimension analogues of the equation for critical vector measures .
again , in the situation of @xcite it was proved that functions @xmath311 are solutions of the cubic equation @xmath312 where @xmath235 is a polynomial and @xmath313 a rational function with poles of order at most @xmath314 .
this fact yields that measures @xmath315 are supported on a finite union of analytic arcs , that are trajectories of a quadratic differential living on the riemann surface of , and which is explicitly given on each sheet of this riemann surface as @xmath316 .
the fact that the support of critical ( or equilibrium ) vector measures is described in terms of trajectories on a compact riemann surface ( and thus , what we actually see is their projection on the plane ) explains the apparent geometric complexity of the limit set for the zeros of hermite pad polynomials , see figure [ fig1hermitepade ] .
now we step into a new territory .
it is well known that a multiplicative modification of an orthogonality measure is easier to handle ( from the point of view of the asymptotic theory ) than an additive one . however
, this kind of problems appears very naturally .
for instance , a large class of transfer functions corresponding to time - delay systems can be written as the ratio of two functions , each of them a polynomial combination of exponentials ( see e.g. @xcite ) . in the simplest case ,
let @xmath317 where @xmath318 and @xmath319 , and we want to consider the best rational approximations of @xmath17 on a given set @xmath320 : @xmath321 .
this is a direct generalization of the central problem solved in @xcite ; in particular , it can be reduced to the asymptotics of polynomials of non - hermitian orthogonality with respect to a varying weight containing @xmath17 , i.e. a _ sum _ of exponentials .
in the rest of the paper we will consider mostly the simplest model situation , which still contains the most essential points , when polynomials @xmath202 are defined by orthogonality relations @xmath322 and the varying ( depending on @xmath4 ) weight function has the form @xmath323 where the integration goes along a curve @xmath324 with @xmath325 being two distinct points fixed on the plane .
let us get ahead of ourselves and announce the status of the problem of the asymptotic zero distribution of polynomials @xmath13 : it depends on the values of the parameters @xmath326 .
in some range the solution is essentially known ; there is another parameters domain where a modification of known methods make it possible to prove a theorem on the zero distribution by generalizing known results . in particular
a new equilibrium problem has to be introduced in order to present these generalizations .
finally , there is a `` gray area '' in the space of parameters , where we can only show some results of numerical experiments and to state partial conjectures explaining those results .
boundaries between above mentioned domains in the space of parameters are not explicitly known yet ; in some cases we have partial results which are mentioned below .
the integral in resembles , at least formally , the orthogonality conditions in , so it is natural to analyze first the applicability of the grs theory , explained in section [ sec : grs ] .
the starting assumption is that the weights @xmath327 in must be analytic in a domain @xmath160 that contains points @xmath328 , where the following limit exists : @xmath329 notice that in the original paper @xcite the convergence in the above formula was assumed to be uniform on compacts in @xmath160 , and consequently the function @xmath191 was harmonic in @xmath160 .
these assumption need to be relaxed for our purposes .
for instance , if in the case we assume @xmath330 , then @xmath331 this function is continuous and piece wise harmonic in @xmath20 ( harmonic in the left @xmath332 and right @xmath333 half - planes ) .
moreover , it is superharmonic and continuous in the whole plane .
convergence in is uniform on compact sets in the complement to the imaginary axis ( notice that @xmath334 is an open set ) .
it is also easy to verify that for an arbitrary compact set in @xmath20 the convergence is still uniform , but from above : for any compact @xmath335 and any @xmath336 there is @xmath337 such that @xmath338 in the light of this example it is natural to try to consider the analogue of the grs theory under the assumptions that the external field @xmath191 is continuous and piece - wise harmonic in @xmath160 , and convergence in is locally uniform in the open sets where @xmath191 is harmonic , and locally uniform from above in @xmath20 . before continuing , and in order to gain some intuition ,
let us present the result of a few numerical experiments , finding the zeros of @xmath13 given by .
these calculations can be carried out in a rather stable way using classical approach via moment determinants , see ( * ? ? ?
* chapter 2 ) .
indeed , @xmath339 where @xmath340 is the incomplete gamma function , so that the moments @xmath341 can be either evaluated in terms of special functions or obtained from the recurrence ( for @xmath342 ) @xmath343 in practice , it might be convenient to apply this recurrence directly to the moments @xmath344 in order to avoid cancellation errors .
using the identities from ( * ? ? ?
* theorem 2.2 ) we find the coefficients of the recurrence relation , satisfied by the polynomials , orthogonal with respect to @xmath345 , and calculate their zeros as eigenvalues of the corresponding ( complex ) jacobi matrix .
all experiments were performed with mathematica version 10.3 , using an extended precision .
let us discuss some of the pictures presented here .
c@c fig7a & fig7b + fig7c & fig7d to start with , we consider the case when both @xmath242 and @xmath243 belong to the same half plane by modifying the orthogonality weight from example [ examplesaffconf ] and taking @xmath346 with the integration in along a jordan arc @xmath214 connecting @xmath239 and @xmath347 , see figure [ figexamplesaffconf ] .
the upper left picture is clear and should be compared with figure [ figexampleexp ] , left : all zeros lie in the left half - plane , where the resulting external field , namely @xmath348 is harmonic .
the asymptotic distribution of zeros is @xmath349 , and @xmath350 is the trajectory of the quadratic differential , joining @xmath242 and @xmath351 .
recall that as @xmath245 increases , this trajectory approaches the imaginary axis , until for certain @xmath352 it touches @xmath353 at a single point . at this point
the @xmath0-property is no longer valid : it does hold for the external field @xmath354 , but not for ! what happens next is shown in figure [ figexamplesaffconf ] , top right .
now , recall that in the standard grs theory the max - min property and the symmetry or @xmath0-property of the support are basically equivalent .
for the piece - wise harmonic @xmath191 this is no longer the case .
indeed , let @xmath153 be the critical trajectory of joining @xmath242 and @xmath351 and that crosses the imaginary axis ( so @xmath355 ) , and let @xmath356 be the continuum obtained by replacing @xmath357 with the straight vertical segment joining the two points of intersection ( see figure [ maxmin ] )
. then @xmath358 ( see the definition in ) .
in other words , at least in the situation of figure [ figexamplesaffconf ] , top right , the zeros of @xmath13 `` prefer '' to follow the max - min configuration instead of the @xmath0-curve , top right , could be formally considered as a limit case of what we describe in theorem [ thm2 ] below , where the @xmath0-property does play a role . nevertheless , the situation is completely different here due to the existence of a string of zeros of the weight along the imaginary axis .
in fact , zeros of @xmath13 , at least visually , are equally spaced on @xmath353 , following the distribution of zeros of @xmath345 and not an equilibrium measure on an interval of @xmath353 . ] .
c@c gamma1 ( 5,32 ) @xmath153 ( 02,135)@xmath359 ( 67,135)@xmath360 & gamma2 ( 5,32 ) @xmath356 ( 02,135)@xmath359 ( 67,135)@xmath360 this does not last too long : the possibility of a symmetry is recovered on figure [ figexamplesaffconf ] , bottom row , which should be compared with figure [ figexampleexp ] .
we observe the appearance of a new connected component of the zero distribution , whose contribution allows for the @xmath0-property of the support , but now in a `` vector '' equilibrium setting , see the discussion next .
c@c fig9a & fig9b in the second set of experiments we consider the case when @xmath361 , that is , @xmath362 in other words , @xmath363 are now determined by the orthogonality conditions @xmath364 in figure [ figcosreal ] we represent the zeros of @xmath365 for @xmath366 ; in particular , we take the totally real case @xmath367 ( figure [ figcosreal ] , left ) , and a slightly perturbed case @xmath368 ( figure [ figcosreal ] , right ) .
notice that all zeros apparently stay either in the left or right half - plane , which corresponds to the situation described by theorem [ thm2 ] below .
c@c fig10a & fig10b however , different values of the parameters @xmath242 and @xmath369 produce different pictures . in figure [ figassym1 ]
we represent the zeros of @xmath251 again for @xmath366 .
but now @xmath370 with @xmath371 ( left ) and @xmath370 with @xmath372 ( right ) .
it looks like the limit distribution of the zeros of @xmath13 can be now either a continuum , cutting the imaginary axis at a single point , or it can split into disjoint arcs , each or some of them also crossing the imaginary axis .
we will give a partial answer ( at least , a conjecture ) for this case in our discussion in the next section .
c@c fig11a & fig11b + fig11c & fig11d let us finally reach the complete symmetry by analyzing the orthogonality , with @xmath373 , figure [ figcosimag ] .
notice that now the path of integration can go entirely along the imaginary axis , precisely where the weight @xmath374 has its zeros .
a similar situation has been analyzed in @xcite for a semi - infinite interval .
we observe that the zeros of @xmath13 in figure [ figcosimag ] can be split basically into three groups , two of them on @xmath353 , and the rest on curves in the complex plane .
each curve belongs entirely ( with a possible exception of its end points ) to a half - plane @xmath375 .
there are also to points @xmath376 , @xmath377 , such that the vertical segment @xmath378 $ ] either contains no zeros of @xmath13 , or they are equally spaced , in concordance with the zeros of the weight @xmath345 there .
the complement of this vertical segment in @xmath379 $ ] also contains zeros of @xmath13 , but now their distribution is quite different , with a clear blow up in the density toward the endpoints @xmath380 .
the whole picture resembles ( but we do not claim at this point anything further beyond a formal similarity ) the constrained equilibrium @xcite , with its void , bands and saturated regions , with the `` constraint '' created by the zeros of the weight .
pictures in figure [ figcosimag ] may look puzzling also in the light of the electrostatic interpretation of the zeros of @xmath13 .
indeed , the cross section of the external field along any line parallel to @xmath152 is convex upward ( i.e. , it looks like the graph of @xmath381 on @xmath152 ) .
such a field should push the zeros away from the imaginary axis , while what we see is more appropriate of a field of the form @xmath382 on @xmath152 . instead of relegating again the explanation of this phenomenon to the next section ,
let us point out here the simple but key fact for the orthogonality , which allows us to `` swap '' the components of the external field .
indeed , the orthogonality condition in can be equivalently rewritten as @xmath383 where @xmath384 , @xmath385 are _ any _ two curves from the class @xmath386 defined in .
alternatively , we could state it as @xmath387 where @xmath22 is an oriented close jordan contour on @xmath20 passing through @xmath242 and @xmath243 , and where @xmath388 is defined either as @xmath389 or @xmath390 on two distinct connected components of @xmath391 .
we see now that we can actually `` pull '' the curve with @xmath392 to the right half - plane ( assuming @xmath245 ) , and the curve with @xmath393 to the left one , so that the corresponding external field becomes just @xmath394 , with @xmath191 in !
so , we have traded a field with the convex upward cross section along @xmath152 for a convex downward one .
the trick we have just performed is an expression of the fact that the orthogonality with respect to a sum of weights can be associated to a vector equilibrium problem , similar to those described in section [ subsec6.7 ] . for more details , see the discussion below .
recall that we are interested in a generalization of the grs theory under the assumptions that the external field @xmath191 is continuous and piece - wise harmonic in @xmath160 , and convergence in is as described at the end of subsection [ subsec6.5 ] .
it turns out that such a generalization is rather straightforward when the support @xmath395 of the equilibrium measure @xmath396 for the @xmath0-curve @xmath0 in the field @xmath191 does not intersect the boundary of the domains where @xmath191 is harmonic . to be more precise
, we assume that : 1 . the sequence @xmath397 for a domain @xmath160 , 2 . there is a finite number of disjoint sub - domains @xmath398 of @xmath399 such that convergence in is uniform on compact subsets of each @xmath400 and locally uniform form above in @xmath160 , 3 .
@xmath191 is continuous in @xmath160 .
observe that by our assumptions , @xmath191 is also harmonic in each @xmath401 .
given a finite set @xmath176 of @xmath177 distinct points in @xmath160 , let @xmath402 denote the family of continua made of a finite union of oriented rectifiable arcs and containing @xmath221 .
[ thm2 ] under the assumptions above , let polynomials @xmath202 be defined by the orthogonality relations , where the integration goes along @xmath403 . if there exists @xmath404 with the @xmath0-property in the external field @xmath191 such that @xmath405 then @xmath406 .
recall that @xmath349 is the equilibrium measure on @xmath0 in the external field @xmath191 and satisfying the symmetry property .
the proof of the original version of the theorem in @xcite is directly applicable to the slightly more general conditions of theorem [ thm2 ] , without any essential modification .
in fact , it is possible to push the situation even further and prove the assertion when @xmath257 cuts the set of discontinuities of @xmath407 in a finite number of points .
let us summarize the most important assumptions of theorem [ thm2 ] : 1 .
the existence of an @xmath0 curve , and 2 .
the fact that @xmath191 is harmonic in a neighborhood of the support of @xmath349 .
it is not an easy problem to find general conditions in terms of the field @xmath191 ( or equivalently , of the weights @xmath345 ) under which both ( a ) and ( b ) are automatically satisfied . in our model case
it is possible at least to indicate a system of equations ( see theorem [ thm-3 ] ) which could lead to a complete or partial solution of the problem . recall the definition of the vector equilibrium from section [ subsec6.7 ] , which involves a vector of @xmath271 nonnegative measures @xmath272 , a symmetric and positive - semidefinite interaction matrix @xmath273 , and a vector of external fields @xmath274 , along with a set of constraints on the size of the components of @xmath294 .
the equilibrium measure is the minimizer of the total vector energy functional .
we call such a problem _ pseudo - vector _ if all entries of the interaction matrix @xmath280 are just @xmath151 s .
the terminology comes from the observation that in this case the total energy is just @xmath408 let us return to our model problem , , that is , @xmath409 where the integration goes along a curve @xmath410 with @xmath325 being two distinct points fixed on the plane , but allowing for a more general form the weight : @xmath411 for simplicity , we assume that @xmath412 and @xmath413 are entire functions , and that @xmath412 converge locally uniformly in @xmath20 ( as @xmath106 ) to a function @xmath414 . as it was observed at the end of section [ sec : numerical experiments ] , conditions are equivalent to @xmath415 where each @xmath416 , and curves @xmath417 can be taken independently from each other .
this reformulation motivates the introduction of the following pseudo - vector problem : for any vector @xmath418 , and the vector of external fields @xmath274 , @xmath419 , harmonic on @xmath20 , consider the energy functional with the additional constraints that @xmath420 , @xmath292 , and @xmath421 as usual , the _ ( pseudo - vector ) equilibrium measure _
@xmath422 will be the minimizer of in this class .
it is characterized by the variational conditions @xmath423 ( compare with ) , where @xmath195 is the equilibrium constant .
the @xmath0-property in the external field @xmath424 has the form now @xmath425 and @xmath194 .
we have seen that equilibrium measures on sets with the @xmath0-property are a subclass of the critical measures , defined by .
this subclass is proper : for each _ critical _ measure @xmath426 we still have @xmath427 but not necessarily @xmath428 moreover , even if is true , we still need to be able to complete the supports of each @xmath429 to the whole curves @xmath417 in such a way that the inequalities in hold in order to claim that @xmath430 . following the arguments of @xcite we see that the variational condition , when considered for functions of the form @xmath431 , @xmath432 , yield the following equation for the cauchy transforms @xmath433 of @xmath429 ,
see the definition in : @xmath434 where @xmath65 is a polynomial .
this identity is valid a.e . in @xmath20 with respect to the plane lebesgue measure .
notice that one of the immediate conclusions is that @xmath435 where @xmath436 , and that each @xmath429 lives on a trajectory of a quadratic differential , but now not necessarily rational . in order to study the structure of such trajectories it would be very convenient to associate with an algebraic equation , in the spirit of what was done in @xcite , which is part of an ongoing project .
the standard grs theorem ( theorem [ grsthm ] ) is directly applicable in the pseudo - vector case when different components of the support of @xmath437 do not intersect ( see , e.g. , figure [ figassym1 ] , right ) : [ thmnew ] under the assumptions above , let polynomials @xmath202 be defined by the orthogonality relations . if there exists a vector @xmath438 , @xmath439 , with the @xmath0-property in the external field @xmath424 , and the corresponding equilibrium measure @xmath440 such that @xmath441 then @xmath442 .
when some components of the support of @xmath437 do intersect , or more precisely , when they contain a common set of positive capacity , the situation is totally open are simple arcs with common endpoints is already interesting . ] .
it looks like in this case the max - min characterization of the support of @xmath437 dominates over its @xmath0-property , but at this stage these are mere speculations .
meanwhile , it is difficult to obtain anything further from , _ unless _ all @xmath443 are linear functions , case we consider next .
the assumption that @xmath191 is piece - wise linear allows for a crucial simplification of the problem , using the observation that after differentiating in all the external fields disappear .
so , let us return to the situation , so that @xmath191 is . the problem contains four parameters : @xmath325 and @xmath444 or six real parameters .
finding the domain in the space of parameters for which an @xmath0-curve does not exist is a formidable task .
it seems more convenient to start with an assumption that the @xmath0-curve does exist and make an ansatz on the structure of the support of its equilibrium measure ; in a piece - wise linear field and curves in the class @xmath445 the support should consist of one or two arcs
. then a straightforward method of proving existence theorems may be used , which we can shortly outline as follows : the analytic representation of the cauchy transform of the equilibrium measure has four unknown complex parameters , constrained by a system of equations expressing the main properties of the problem . using any solution of such a system one can prove existence of @xmath0-curve by direct construction . in a similar situation
the method has been used in @xcite . for the orthogonality and the corresponding external field this approach yields the following result : [ thm-3 ] assume that @xmath0 is an @xmath0-curve in the external field , that the support of the equilibrium measure @xmath446 does not intersect the imaginary axis and that it has at most two connected components
. then the following assertions are valid : * the are two quadratic polynomials , @xmath447 such that the cauchy transform of the equilibrium measure @xmath448 has the form @xmath449 where @xmath450 , and the branch of the root in the denominator is fixed in @xmath451 by the condition @xmath452 . *
the parameters @xmath453 , @xmath454 , @xmath455 , @xmath456 are determined by the system of equations @xmath457 * furthermore , @xmath458 where @xmath459 is a critical trajectory of the quadratic differential @xmath460 , with @xmath461 function @xmath462 is holomorphic in the left half plane @xmath463 , while @xmath464 is holomorphic in @xmath465 .
it follows the lines of @xcite .
the @xmath0-property , the fact that @xmath466 and the variational equation imply the general form of @xmath467 .
equation means that @xmath468 which follows again from .
equation is equivalent to @xmath469 which is a necessary condition for the existence of trajectories joining @xmath258 with @xmath470 , while is just an expression of the fact that @xmath110 is a unit measure .
finally , imposes the equality of the equilibrium constants , .
the existence of the @xmath0-curve in the assumptions of theorem [ thm-3 ] also implies that arcs of @xmath257 may be completed to a curve from @xmath210 preserving the inequalities in . on the other hands ,
the existence of a solution of does not guarantee a priori the existence of the necessary critical trajectories of @xmath460 . in our case we can prove it using the homotopy arguments , starting with the totally symmetric case , analyzed next .
assume that @xmath471 ( see figure [ figcosreal ] , left , with @xmath472 and @xmath473 ) , so that @xmath349 its just the equilibrium measure of the segment @xmath474 $ ] in the external field @xmath475 , and its existence and uniqueness is guaranteed by the classical theory , see e.g. @xcite . as usual ,
the main step consists in finding the support @xmath476 of @xmath193 , which in this case , for any values of the parameters @xmath477 and @xmath369 in , takes the form @xmath478 \cup [ v , a ] , \quad\text { where}\quad v = v(a , k ) \in ( 0,a ) ; \ ] ] the fact that @xmath479 follow from @xcite , as well as from the calculations below .
once the support is determined , we can find the expression for the cauchy transform @xmath467 of @xmath110 , and from there to recover the equilibrium measure ( and by theorem [ thm2 ] , the asymptotics of @xmath480 ) by sokhotsky
plemelj s formula .
notice that the @xmath0-property is satisfied simply by symmetry reasons , so the existence of an @xmath0-curve for any such @xmath481 is now automatic .
the assertion of theorem [ thm-3 ] boils down now to @xmath482 where @xmath483 and the branch of @xmath484 in @xmath485 is positive for @xmath486 . from these expressions we get @xmath110 , which is absolutely continuous on @xmath395 : @xmath487,\ ] ] where as usual , @xmath488 are the boundary values of this function on @xmath489 form the upper half plane .
the two real parameters @xmath490 and @xmath491 in are defined by the equations @xmath492 the first identity expressing the equality of the equilibrium constants on both components of @xmath395 , and the second one assuring that @xmath110 is a probability measure , with its mass distributed evenly on @xmath493 $ ] and @xmath494 $ ] . for every @xmath495 and @xmath496
there is a unique solution @xmath497 of this system , and hence , a unique solution of the @xmath0-problem for this case .
now assume that @xmath498 we can no longer guarantee the existence of an @xmath0-curve in class @xmath499 of curves connecting @xmath242 and @xmath243 for all values of the parameters @xmath481 . however , it is intuitively clear that for @xmath500 small enough , an @xmath0 curve exists , is discontinuous , and it still lies in the domain where @xmath191 is harmonic ( cf . figure [ figcosreal ] , right , with @xmath501 and @xmath473 ) .
it turns out that the formulas , and remain valid with obvious modifications and equations for the parameters are similar to those in . for other values of the parameters
the limit distribution @xmath110 of the zeros of @xmath13 can be either a continuum , cutting the imaginary axis in a single ( or a finite number of ) point ( cf . figure [ figassym1 ] , left , with @xmath370 and @xmath371 ) , or it can split into disjoint arcs , each or some of them crossing the imaginary axis ( cf . figure [ figassym1 ] , right , with @xmath370 and @xmath372 ) . in this case
we expect these arcs to be the support of the equilibrium measure for the pseudo - vector problem described in section [ sec : pseudo ] .
we conclude our exposition with the remark that the reformulation of the orthogonality conditions as could be crucial for the application of the nonlinear steepest descent analysis of deift - zhou @xcite for the study of the strong ( and in consequence , also weak- * ) asymptotics of @xmath13 s , especially in the most elusive cases when the @xmath0-property is apparently lost .
we plan to address this problem in a forthcoming publication .
this paper is based on a plenary talk at the focm conference in montevideo , uruguay , in 2014 by the first author ( amf ) , who is very grateful to the organizing committee for the invitation .
this work was completed during a stay of amf as a visiting chair professor at the department of mathematics of the shanghai jiao tong university ( sjtu ) , china .
he acknowledges the hospitality of the host department and of the sjtu .
amf was partially supported by the spanish government together with the european regional development fund ( erdf ) under grants mtm2011 - 28952-c02 - 01 ( from micinn ) and mtm2014 - 53963-p ( from mineco ) , by junta de andaluca ( the excellence grant p11-fqm-7276 and the research group fqm-229 ) , and by campus de excelencia internacional del mar ( ceimar ) of the university of almera .
we are grateful to s. p. suetin for allowing us to use the results of his calculations in figure [ fig1hermitepade ] , as well as to an anonymous referee for several remarks that helped to improve the presentation .
a. i. aptekarev and h. stahl .
asymptotics of hermite - pad polynomials . in _
progress in approximation theory ( tampa , fl , 1990 ) _ , volume 19 of _ springer ser .
_ , pages 127167 .
springer , new york , 1992 .
j. baik , t. kriecherbauer , k. d. t .- r .
mclaughlin , and p. d. miller .
uniform asymptotics for polynomials orthogonal with respect to a general class of discrete weights and universality results for associated ensembles : announcement of results . , ( 15):821858 , 2003 .
a. a. gonchar and e. a. rakhmanov . on the convergence of simultaneous pad approximants for systems of functions of markov type . , 157:3148 , 234 , 1981 .
number theory , mathematical analysis and their applications .
a. a. gonchar and e. a. rakhmanov .
equilibrium measure and the distribution of zeros of extremal polynomials .
, 125(2):117127 , 1984 .
translation from mat .
134(176 ) , no.3(11 ) , 306 - 352 ( 1987 ) . a. a. gonchar and e. a. rakhmanov .
equilibrium distributions and degree of rational approximation of analytic functions . , 62(2):305348 , 1987 .
translation from mat .
134(176 ) , no.3(11 ) , 306 - 352 ( 1987 ) .
g. lpez lagomasino and a. ribalta .
approximation of transfer functions of unstable infinite dimensional control systems by rational interpolants with prescribed poles . in
_ proceedings of the international conference on rational approximation , icra99 ( antwerp ) _ , volume 61 , pages 267294 , 2000 .
a. martines - finkelshten , e. a. rakhmanov , and s. p. suetin .
differential equations for hermite - pad polynomials . , 68(1(409)):197198 , 2013 .
translation in russian math .
surveys 68 ( 2013 ) , no .
1 , 183185 . a. martnez - finkelshtein , p. martnez - gonzlez , and r. orive .
zeros of jacobi polynomials with varying non - classical parameters . in _ special functions ( hong kong , 1999 ) _ , pages 98113 .
world sci .
publ . , river edge , nj , 2000 .
a. martnez - finkelshtein , p. martnez - gonzlez , and f. thabet .
trajectories of quadratic differentials for jacobi polynomials with complex parameters .
preprint arxiv math.1506.03434 , to appear in comput .
methods and funct .
theory , 2015 .
a. martnez - finkelshtein and e. a. rakhmanov . on asymptotic behavior of heine - stieltjes and van vleck polynomials . in _ recent trends in orthogonal polynomials and
approximation theory _ ,
volume 507 of _ contemp .
_ , pages 209232 .
soc . , providence , ri , 2010 .
a. martnez - finkelshtein , e. a. rakhmanov , and s. p. suetin .
heine , hilbert , pad , riemann , and stieltjes : john nuttall s work 25 years later . in _ recent advances in orthogonal polynomials , special functions , and their applications _ ,
volume 578 of _ contemp .
_ , pages 165193 .
amer . math .
soc . , providence , ri , 2012 .
j. nuttall .
the convergence of pad approximants to functions with branch points . in _ pad and rational approximation ( proc .
south florida , tampa , fla .
, 1976 ) _ , pages 101109 . academic press , new york , 1977 .
frank w. j. olver , daniel w. lozier , ronald f. boisvert , and charles w. clark , editors . .
department of commerce , national institute of standards and technology , washington , dc ; cambridge university press , cambridge , 2010 . with 1 cd - rom ( windows , macintosh and unix ) .
e. a. rakhmanov .
orthogonal polynomials and @xmath0-curves . in _ recent advances in orthogonal polynomials , special functions , and their applications _ ,
volume 578 of _ contemp .
_ , pages 195239 .
soc . , providence , ri , 2012 .
e. b. saff , j. l. ullman , and r. s. varga .
incomplete polynomials : an electrostatics approach . in _ approximation theory , iii ( proc . conf .
, univ . texas , austin , tex . , 1980 )
_ , pages 769782 . academic press , new york , 1980 . h. stahl . on the divergence of certain pad approximant and the behaviour of the associated orthogonal polynomials . in _
orthogonal polynomials and applications ( bar - le - duc , 1984 ) _ , volume 1171 of _ lecture notes in math .
_ , pages 321330 .
springer , berlin , 1985 .
h. stahl .
asymptotics of hermite - pad polynomials and related convergence results a summary of results . in _ nonlinear numerical methods and rational approximation ( wilrijk , 1987 )
_ , volume 43 of _ math .
_ , pages 2353 .
reidel , dordrecht , 1988 . | the complex or non - hermitian orthogonal polynomials with analytic weights are ubiquitous in several areas such as approximation theory , random matrix models , theoretical physics and in numerical analysis , to mention a few . due to the freedom in the choice of the integration contour for such polynomials , the location of their zeros is a priori not clear .
nevertheless , numerical experiments , such as those presented in this paper , show that the zeros not simply cluster somewhere on the plane , but persistently choose to align on certain curves , and in a very regular fashion .
the problem of the limit zero distribution for the non - hermitian orthogonal polynomials is one of the central aspects of their theory .
several important results in this direction have been obtained , especially in the last 30 years , and describing them is one of the goals of the first parts of this paper .
however , the general theory is far from being complete , and many natural questions remain unanswered or have only a partial explanation .
thus , the second motivation of this paper is to discuss some `` mysterious '' configurations of zeros of polynomials , defined by an orthogonality condition with respect to a sum of exponential functions on the plane , that appeared as a results of our numerical experiments . in this apparently simple situation
the zeros of these orthogonal polynomials may exhibit different behaviors : for some of them we state the rigorous results , while others are presented as conjectures ( apparently , within a reach of modern techniques ) .
finally , there are cases for which it is not yet clear how to explain our numerical results , and where we can not go beyond an empirical discussion . |
we will work up to the general question by first examining the special ( low ratings ) case when one candidate has at least as many votes as the other throughout the tally .
this is the classical `` ballot problem '' , in which candidate e and candidate n are competing for a public office .
candidate e wins the election with @xmath0 votes .
how many ways are there to report the votes so that at all times during the tally n is not ahead of e ?
we may represent the state of the tally at any moment by the pair @xmath1 , where the coordinates @xmath2 and @xmath3 count the votes received by e and n respectively .
then a tally consists of a sequence of points on the integer lattice in the plane made in steps of @xmath4 and @xmath5 .
such a sequence is called a _ northeast lattice
path_. we say that the lattice path @xmath6 is _ restricted _ by the lattice path @xmath7 if no part of @xmath6 lies directly above @xmath7 .
for example , figure [ restrictedpaths ] shows two northeast lattice paths from @xmath8 to @xmath9 that are restricted by the `` staircase '' @xmath10 , or , equivalently , that do not go above the line @xmath11 .
the ballot problem asks for the number @xmath12 of these paths .
( note that if the tally ends at @xmath13 , we may uniquely continue it to a northeast lattice path ending at @xmath9 . )
[ restrictedpaths ] to @xmath14 restricted by @xmath15 . ]
the ballot problem can be solved by constructing a simple recurrence .
let @xmath6 be a northeast lattice path restricted by the staircase @xmath7 .
consider the point on @xmath6 where it first revisits the line @xmath16 , and let @xmath17 be the @xmath2-coordinate of this point .
( this point exists since @xmath6 ends at @xmath9 . ) for the upper path in figure [ restrictedpaths ] , @xmath18 ; for the lower path , @xmath19 .
notice that since @xmath6 does not go above @xmath16 and begins at @xmath8 its first step is @xmath20 ; further , its last step before reaching the point @xmath21 is @xmath22 .
therefore we may delete these steps to obtain a northeast lattice path from @xmath23 to @xmath24 that does not go above the line @xmath25 .
there are @xmath26 ways to form such a path , and there are @xmath27 ways to continue this path from @xmath21 to @xmath9 , so we have that @xmath28 .
this we recognize as the familiar recurrence satisfied by the catalan numbers @xmath29 ( * ? ? ?
* exercise 6.19(h ) ) , so we simply check that the initial condition @xmath30 agrees .
we now consider a generalization of the ballot problem .
let @xmath31 be the number of northeast lattice paths restricted by an arbitrary northeast lattice path @xmath7 from @xmath8 to @xmath13 .
the path @xmath7 represents the network s predetermined restrictions on the tally .
it was known by macmahon @xcite that the sum of @xmath31 over all such paths is @xmath32 however , we are interested in computing @xmath31 for specific @xmath7 .
first we develop notation for lattice paths .
it is possible to represent a northeast lattice path as a word on @xmath33 , such as @xmath34 for the upper path in figure [ restrictedpaths ] .
however , this representation is redundant , because the location of each @xmath20 step determines the path uniquely .
therefore , we may represent a northeast lattice path by the sequence of heights @xmath35 of the path along each interval from @xmath36 to @xmath37 .
for example , for the upper path in figure [ restrictedpaths ] we have @xmath38 .
this representation is always a nondecreasing tuple of integers , and it is our primary representation of lattice paths in this note .
a lattice path @xmath39 is restricted by the lattice path @xmath40 precisely when @xmath41 componentwise , i.e. , @xmath42 whenever @xmath43 . to write the main result
, however , it turns out to be more natural to use still another representation of a northeast lattice path @xmath7 its difference sequence @xmath44 let @xmath45 .
since @xmath7 is a northeast lattice path , the entries of @xmath46 are nonnegative integers .
the entry @xmath47 is the number of @xmath22 steps taken along the line @xmath48 , so we can think of this representation as determining a path by the location of each @xmath22 step .
the operator @xmath49 has an inverse @xmath50 , which produces the sequence of partial sums : @xmath51 the relationship between @xmath7 and @xmath52 can be interpreted in another way . if @xmath53 is a tuple of nonnegative integers , the pitman
stanley polytope @xcite defined by @xmath46 is @xmath54 thus a tuple @xmath55 of nonnegative integers is a lattice point inside @xmath56 precisely when the northeast lattice path @xmath57 is restricted by @xmath58 . in other words
, @xmath49 provides a bijection from the northeast lattice paths restricted by @xmath7 to the lattice points in @xmath59 .
we now return to the question at hand : how many northeast lattice paths are restricted by the path @xmath60 ?
equivalently , how many lattice points lie inside @xmath59 ?
one answer to this question is the following determinant enumeration .
let @xmath61 be the @xmath62 matrix with entries @xmath63 .
then the number of northeast lattice paths restricted by @xmath7 is @xmath64 , as given by kreweras @xcite and mohanty ( * ? ? ? * theorem 2.1 ) .
this fact can be obtained from the triangular system of equations @xmath65 for @xmath66 ( where @xmath67 ) , which comes from an inclusion exclusion argument ; solve for @xmath68 using cramer s rule , and in the numerator expand by minors along the last column .
the following theorem presents a formula for @xmath31 in which the lattice points in @xmath69 play a central role .
this gives a non - determinantal formula for the number of northeast lattice paths restricted by @xmath7 .
a generalization of the formula has been independently discovered by gessel and by pitman and stanley ( * ? ? ?
* equation ( 33 ) ) in more advanced contexts .
our proof uses elementary combinatorial methods .
let @xmath7 be a northeast lattice path from @xmath8 to @xmath13 , and let @xmath52 .
the number of northeast lattice paths restricted by @xmath7 is @xmath70 where the sum is over all @xmath71 lattice points @xmath2 in @xmath69 .
we immediately obtain two well - known results as special cases .
for @xmath72 we see that @xmath73 , which gives @xmath74 since @xmath75 ( from the generalization of the binomial theorem @xmath76 to @xmath77 ) , the only nonzero terms in the sum come from lattice points of the form @xmath78 , and therefore @xmath79 as expected . for @xmath80
we recover the ballot problem .
namely , @xmath81 , so @xmath82 equation ( [ thm ] ) allows one to compute @xmath31 not only for explicit integer paths but for symbolic paths , and the resulting expressions have the pleasant property that they are written in the basis of rising factorials @xmath83 .
for example , @xmath84 . for a general path of length @xmath85
, we have @xmath86 and @xmath87 is @xmath88 putting equation ( [ thm ] ) together with the determinantal formula for @xmath31 , we obtain a formula for a certain symbolic determinant in the same basis : @xmath89 where again @xmath52 .
we note that amdeberhan and stanley ( * ? ? ?
* corollary 4.7 ) show that @xmath31 also gives the number of monomials in the expanded form of the multivariate polynomial @xmath90 in the variables @xmath91 .
moreover , @xmath31 is the number of noncrossing matchings of a certain type ( * ? ? ?
* corollary 4.9 ) .
let @xmath92 be the number of lattice points in @xmath56 , where @xmath93 .
that is , @xmath94 . the following recurrence will be used .
we have @xmath95 the only lattice point in @xmath96 is @xmath97 ; hence @xmath98 . for @xmath99 , we partition the lattice points @xmath100 in @xmath56 according to the first entry @xmath101 .
since @xmath100 is a lattice point in @xmath56 , then @xmath102 , so @xmath103 .
therefore , lattice points @xmath104 in @xmath56 are in bijection ( by deleting the first entry @xmath105 ) with lattice points in @xmath106 .
thus @xmath107 is the number of lattice points in @xmath56 with first entry @xmath105 , giving the recurrence . to prove the theorem , then , it suffices to show that equation ( [ thm ] ) satisfies this recurrence .
the base case @xmath108 is easily checked , since the product is empty ; we have @xmath109 since again @xmath96 has only one lattice point .
the remainder of this note is devoted to showing that for @xmath99 @xmath110 where the left sum is over all @xmath71 lattice points @xmath2 in @xmath69 and the right sum is over all @xmath111 lattice points @xmath3 in @xmath112 .
we proceed by simplifying this equation until it becomes a statement about sums of binomial coefficients , given in the lemma below .
first interchange the two summations on the right side of equation ( [ rec ] ) .
next , fix @xmath113 on the right side , and break up the sum on the left according to the choice of @xmath3 in the following way .
the _ children _ of @xmath113 are the elements of the set @xmath114 for example , the children of the lattice point @xmath115 are @xmath116 , @xmath117 , @xmath118 , and @xmath119 .
it is immediate that each lattice point @xmath2 has a unique parent @xmath3 .
this definition is central to the proof .
the reason for defining children in this way is that @xmath2 is a lattice point in @xmath120 if and only if @xmath2 s parent is a lattice point in @xmath112 .
this property provides a many - to - one correspondence between the @xmath0-dimensional lattice points in @xmath69 and the ( @xmath121)-dimensional lattice points in @xmath112 . using this correspondence to break up equation ( [ rec ] ) ,
we obtain @xmath122 for each @xmath3 , where the left sum is over all children @xmath123 of @xmath3 .
it now suffices to prove equation ( [ eq2 ] ) for a fixed @xmath3 , since summing both sides of equation ( [ eq2 ] ) over all @xmath12 lattice points @xmath3 in @xmath112 produces equation ( [ rec ] ) .
note that if @xmath2 is a child of @xmath3 then @xmath124 for @xmath125 , so we may divide both sides of equation ( [ eq2 ] ) by the product @xmath126 to obtain @xmath127 we know what the children of @xmath3 look like , so the sum on the left side can be written as @xmath128 the first term in this expression , which corresponds to the child @xmath129 of @xmath3 , is equal to the @xmath130 term on the right side of equation ( [ eq3 ] ) . removing this term from both sides leaves
@xmath131 which is proved in the following lemma under the substitution @xmath132 , @xmath133 , and @xmath134 .
let @xmath135 , @xmath136 , and @xmath137 be nonnegative integers .
then @xmath138 we show that both sides of the equation are equal to @xmath139 the right side is a telescoping sum : @xmath140 the result for the left side follows from a generalization of the vandermonde identity , namely @xmath141 ( * ? ? ?
* problem 1.42(i ) ) .
the summand on the left side of this equation counts the @xmath142-element subsets of @xmath143 whose @xmath144st element is @xmath145 by choosing @xmath146 of the first @xmath147 elements to be _ not _ in the set and @xmath148 of the last @xmath149 elements to be _ not _ in the set .
the right side counts all @xmath142-element subsets of @xmath143 by selecting the elements _ not _ in the set .
subtract @xmath150 from both sides of this equation and substitute @xmath151 , @xmath152 , @xmath153 , and @xmath154 to obtain @xmath155 thus the director of programming may , for example , determine the likelihood that a random tally of votes will satisfy the network s needs .
we thank dimitrije kostic for generating our interest in this topic @xcite , tewodros amdeberhan for pointing us to relevant literature , doron zeilberger for encouraging us to simplify , and an anonymous referee for a number of helpful suggestions . | we provide an elementary proof of a formula for the number of northeast lattice paths that lie in a certain region of the plane .
equivalently , this formula counts the lattice points inside the pitman
stanley polytope of an @xmath0-tuple .
suppose that on election day a tv news network of questionable morality wants to increase their viewership as polling results come in .
while the reporters can not control the outcome of the election , they can control the order in which votes are reported to the public .
if one candidate is ahead in the tally throughout the entire day , viewership will wane since it is clear that she will win the election . on the other hand ,
a more riveting broadcast occurs when one candidate is ahead at certain times and the other candidate is ahead at others .
in fact , the network employs a group of psychologists and market analysts who have worked out certain margins they would like to achieve at certain points in the tally .
the director of programming needs to know the number of ways this can be done . |
by exploiting signal sparsity and smart reconstruction schemes , compressive sensing ( cs ) @xcite can enable signal acquisition with fewer measurements than traditional sampling . in cs ,
an @xmath0-dimensional signal @xmath1 is measured through @xmath2 random linear measurements .
although the signal may be undersampled ( @xmath3 ) , it may be possible to recover @xmath1 assuming some sparsity structure .
so far , most of the cs literature has considered signal recovery directly from linear measurements . however , in many practical applications , measurements have to be discretized to a finite number of bits .
the effect of such quantized measurements on the performance of the cs reconstruction has been studied in @xcite . in @xcite
the authors adapt cs reconstruction algorithms to mitigate quantization effects . in @xcite ,
high - resolution functional scalar quantization theory was used to design quantizers for lasso reconstruction @xcite .
the contribution of this paper to the quantized cs problem is twofold : first , for quantized measurements , we propose reconstruction algorithms based on gaussian approximations of belief propagation ( bp ) .
bp is a graphical model - based estimation algorithm widely used in machine learning and channel coding @xcite that has also received significant recent attention in compressed sensing @xcite .
although exact implementation of bp for dense measurement matrices is generally computationally difficult , gaussian approximations of bp have been effective in a range of applications @xcite .
we consider a recently developed gaussian - approximated bp algorithm , called _ relaxed belief propagation _
@xcite , that extends earlier methods @xcite to nonlinear output channels .
we show that the relaxed bp method is computationally simple and , with quantized measurements , provides significantly improved performance over traditional cs reconstruction based on convex relaxations .
our second contribution concerns the quantizer design . with linear reconstruction and mean - squared error distortion
, the optimal quantizer simply minimizes the mean squared error ( mse ) of the transform outputs .
thus , the quantizer can be optimized independently of the reconstruction method .
however , when the quantizer outputs are used as an input to a nonlinear estimation algorithm , minimizing the mse between quantizer input and output is not necessarily equivalent to minimizing the mse of the final reconstruction . to optimize the quantizer for the relaxed bp algorithm , we use the fact that the mse under large random transforms can be predicted accurately from a set of simple state evolution ( se ) equations @xcite
then , by modeling the quantizer as a part of the measurement channel , we use the se formalism to optimize the quantizer to asymptotically minimize distortions after the reconstruction by relaxed bp .
in a noiseless cs setting the signal @xmath4 is acquired via @xmath3 linear measurements of the type @xmath5 where @xmath6 is the _ measurement matrix_. the objective is to recover @xmath1 from @xmath7 . although the system of equations formed is underdetermined , the signal is still recoverable if some favorable assumptions about the structure of @xmath1 and @xmath8 are made .
generally , in cs the common assumption is that the signal is exactly or approximately sparse in some orthonormal basis @xmath9 , i.e. , there is a vector @xmath10 with most of its elements equal or close to zero .
additionally , for certain guarantees on the recoverability of the signal to hold , the matrix @xmath8 must satisfy the _ restricted isometry property _ ( rip ) @xcite .
some families of random matrices , like appropriately - dimensioned matrices with i.i.d .
gaussian elements , have been demonstrated to satisfy the rip with high probability . a common method for recovering the signal from the measurements is basis pursuit .
this involves solving the following optimization problem : @xmath11 where @xmath12 is the @xmath13-norm of the signal .
although it is possible to solve basis pursuit in polynomial time by casting it as a linear program ( lp ) @xcite , its complexity has motivated researchers to look for even cheaper alternatives like numerous recently - proposed iterative methods @xcite . moreover , in real applications cs reconstruction scheme must be able to mitigate imperfect measurements , due to noise or limited precision @xcite . a quantizer is a function that discretizes its input by performing a mapping from a continuous set to some discrete set . more specifically , consider @xmath14-point regular scalar quantizer @xmath15 , defined by its output levels @xmath16 , decision boundaries @xmath17 , and a mapping @xmath18 when @xmath19 @xcite .
additionally define the inverse image of the output level @xmath20 under @xmath15 as a _ cell _
for @xmath22 , if @xmath23 we replace the closed interval @xmath24 by an open interval @xmath25 .
typically quantizers are optimized by selecting decision boundaries and output levels in order to minimize the distortion between the random vector @xmath26 and its quantized representation @xmath27 .
for example , for a given vector @xmath28 and the mse distortion metric , optimization is performed by solving @xmath29 where minimization is done over all @xmath14-level regular scalar quantizers .
one standard way of optimizing @xmath15 is via the _
lloyd algorithm _ , which iteratively updates the decision boundaries and output levels by applying necessary conditions for quantizer optimality @xcite .
however , for the cs framework finding the quantizer that minimizes mse between @xmath28 and @xmath30 is not necessarily equivalent to minimizing mse between the sparse vector @xmath31 and its cs reconstruction from quantized measurements @xmath32 @xcite .
this is due to the nonlinear effect added by any particular cs reconstruction function .
hence , instead of solving ( [ equ : quantization : optimal ] ) , it is more interesting to solve @xmath33 where minimization is performed over all @xmath14-level regular scalar quantizers and @xmath32 is obtained through a cs reconstruction method like relaxed bp or amp .
this is the approach taken in this work .
consider the problem of estimating a random vector @xmath34 from noisy measurements @xmath35 , where the noise is described by a measurement channel @xmath36 , which acts identically on each measurement @xmath37 of the vector @xmath38 obtained via ( [ equ : cs : measurement ] ) .
moreover suppose that elements in the vector @xmath31 are distributed i.i.d . according to @xmath39 .
then we can construct the following conditional probability distribution over random vector @xmath31 given the measurements @xmath40 : @xmath41 where @xmath42 is the normalization constant and @xmath43 . by marginalizing this distribution
it is possible to estimate each @xmath44 .
although direct marginalization of @xmath45 is computationally intractable in practice , we approximate marginals through bp @xcite .
bp is an iterative algorithm commonly used for decoding of ldpc codes @xcite .
we apply bp by constructing a bipartite factor graph @xmath46 from ( [ equ : rbp : marginalization ] ) and passing the following messages along the edges @xmath47 of the graph : @xmath48 where @xmath49 means that the distribution is to be normalized so that it has unit integral and integration is over all the elements of @xmath1 except @xmath44 .
we refer to messages @xmath50 as variable updates and to messages @xmath51 as measurement updates .
we initialize bp by setting @xmath52 .
earlier works on bp reconstruction have shown that it is asymptotically mse optimal under certain verifiable conditions .
these conditions involve simple single - dimensional recursive equations called _ state evolution ( se ) _ , which predicts that bp is optimal when the corresponding se admits a unique fixed point @xcite .
nonetheless , direct implementation of bp is still impractical due to the dense structure of @xmath8 , which implies that the algorithm must compute the marginal of a high - dimensional distribution at each measurement node .
however , as mentioned in section [ sec : intro ] , bp can be simplified through various gaussian approximations , including the _ relaxed bp _ method @xcite and _ approximate message passing ( amp )
_ @xcite .
recent theoretical work and extensive numerical experiments have demonstrated that , in the case of certain large random measurement matrices , the error performance of both relaxed bp and amp can also be accurately predicted by se . hence the optimal quantizers can be obtained in parallel for both of the methods , however in this paper we concentrate on design for relaxed bp , while keeping in mind that identical work can be done for amp as well .
due to space limitations , in this paper we will limit our presentation of relaxed bp and se equations to the setting in figure [ fig : probmod ] .
see @xcite for more general and detailed analysis .
the vector @xmath53 denotes noiseless random measurements.,width=307 ] consider the cs setting in figure [ fig : probmod ] , where without loss of generality we assumed that @xmath54 .
the vector @xmath34 is measured through the random matrix @xmath55 to result in @xmath56 , which is further perturbed by some additive white gaussian noise ( awgn ) .
the resulting vector @xmath28 can be written as @xmath57 where @xmath58 are i.i.d .
random variables distributed as @xmath59 .
these noisy measurements are then quantized by the @xmath14-level scalar quantizer @xmath15 to give the cs measurements @xmath35 .
the relaxed bp algorithm is used to estimate the signal @xmath31 from the corrupted measurements @xmath40 , given the matrix @xmath55 , noise variance @xmath60 , and the quantizer mapping @xmath15 . note that under this model each quantized measurement @xmath61 indicates that @xmath62
, hence our measurement channel can be characterized as @xmath63 for @xmath64 and where @xmath65 is gaussian function @xmath66 relaxed bp can be implemented by replacing probability densities in ( [ equ : rbp : varupdatebp ] ) and ( [ equ : rbp : mesupdatebp ] ) by two scalar parameters each , which can be computed according to the following rules : @xmath67 where @xmath68 is the variance of the components @xmath69 .
additionally , at each iteration we estimate the signal via @xmath70 for each @xmath71 .
we refer to messages @xmath72 as variable updates and to messages @xmath73 as measurement updates .
the algorithm is initialized by setting @xmath74 and @xmath75 where @xmath76 and @xmath77 are the mean and variance of the prior @xmath78 .
the nonlinear functions @xmath79 and @xmath80 are the conditional mean and variance @xmath81 where @xmath82 , @xmath83 , and @xmath84 .
note that these functions admit closed - form expressions and can easily be evaluated for the given values of @xmath85 and @xmath86 .
similarly , the functions @xmath87 and @xmath88 can be computed via @xmath89 where the functions @xmath90 and @xmath91 are the conditional mean and variance @xmath92 of the random variable @xmath93 .
these functions admit closed - form expressions in terms of @xmath94 .
the equations ( [ equ : rbp : varupdaterbpmean])([equ : rbp : varestimationrbp ] ) are easy to implement , however they provide us no insight into the performance of the algorithm .
the goal of se equations is to describe the asymptotic behavior of relaxed bp under large measurement matrices .
the se for our setting in figure [ fig : probmod ] is given by the recursion @xmath95 where @xmath96 is the iteration number , @xmath97 is a fixed number denoting the measurement ratio , and @xmath68 is the variance of the awgn components which is also fixed .
we initialize the recursion by setting @xmath98 , where @xmath77 is the variance of @xmath44 according to the prior @xmath78 .
we define the function @xmath99 as @xmath100 where the expectation is taken over the scalar random variable @xmath82 , with @xmath101 , and @xmath102 .
similarly , the function @xmath103 is defined as @xmath104 where @xmath88 is given by ( [ equ : rbp : d2 ] ) and the expectation is taken over @xmath36 and @xmath105 , with the covariance matrix @xmath106 one of the main results of @xcite , which we present below for completeness , was to demonstrate the convergence of the error performance of the relaxed bp algorithm to the se equations under large sparse measurement matrices .
denote by @xmath107 the number of nonzero elements per column of @xmath8 . in the large sparse limit analysis , first let @xmath108 with @xmath109 and keeping @xmath110 fixed .
this enables the local - tree properties of the factor graph @xmath111 .
then let @xmath112 , which will enable the use of a central limit theorem approximation .
[ thm : rbptoseconvergence ] consider the relaxed bp algorithm under the large sparse limit model above with transform matrix @xmath8 and index @xmath113 satisfying the assumption 1 of @xcite for some fixed iteration number @xmath114 .
then the error variances satisfy the limit @xmath115 where @xmath116 is the output of the se equation ( [ equ : se : serecursion ] ) .
see @xcite .
another important result regarding se recursion in ( [ equ : se : serecursion ] ) is that it admits at least one fixed point .
it has been showed that as @xmath117 the recursion decreases monotonically to its largest fixed point and if the se admits a unique fixed point , then relaxed bp is asymptotically mean - square optimal @xcite .
although in practice measurement matrices are rarely sparse , simulations show that se predicts well the behavior of relaxed bp .
moreover , recently more sophisticated techniques were used to demonstrate the convergence of approximate message passing algorithms to se under large i.i.d .
gaussian matrices @xcite .
we now return to the problem of designing mse - optimal quantizers under relaxed bp presented in ( [ equ : quantization : csoptimal ] ) . by modeling the quantizer as part of the channel and working out the resulting equations for relaxed bp and se
, we can make use of the convergence results to recast our optimization problem to @xmath118 where minimization is done over all @xmath14-level regular scalar quantizers . in practice ,
about 10 to 20 iterations are sufficient to reach the fixed point of @xmath116 . then by applying theorem [ thm : rbptoseconvergence ] , we know that the asymptotic performance of @xmath119 will be identical to that of @xmath120 .
it is important to note that the se recursion behaves well under quantizer optimization .
this is due to the fact that se is independent of actual output levels and small changes in the quantizer boundaries result in only minor change in the recursion ( see ( [ equ : rbp : eout ] ) ) .
although closed - form expressions for the derivatives of @xmath116 for large @xmath114 s are difficult to obtain , we can approximate them by using finite difference methods .
finally , the recursion itself is fast to evaluate , which makes the scheme in ( [ equ : quantization : seoptimal ] ) practically realizable under standard optimization methods like sequential quadratic programming ( sqp ) .
we now present experimental validation for our results .
assume that the signal @xmath31 is generated with i.i.d .
elements from the gauss - bernoulli distribution @xmath121 where @xmath122 is the sparsity ratio that represents the average fraction of nonzero components of @xmath31 . in the following experiments we assume @xmath123 .
we form the measurement matrix @xmath8 from i.i.d .
gaussian random variables , i.e. , @xmath124 ; and we assume that awgn with variance @xmath125 perturbs measurements before quantization .
now , we can formulate the se equation ( [ equ : se : serecursion ] ) and perform optimization ( [ equ : quantization : seoptimal ] ) .
we compare two cs - optimized quantizers : _ uniform _ and _ optimal_. we fix boundary points @xmath23 and @xmath126 , and compute the former quantizer through optimization of type ( [ equ : quantization : optimal ] ) . in particular , by applying the central limit theorem we approximate elements @xmath127 of @xmath28 to be gaussian and determine the _ uniform _ quantizer by solving ( [ equ : quantization : optimal ] ) , but with an additional constraint of equally - spaced output levels . to determine _ optimal _ quantizer , we perform ( [ equ : quantization : seoptimal ] ) by using a standard sqp optimization algorithm for nonlinear continuous optimization . bits / component of @xmath31 .
optimal quantizer is found by optimizing quantizer boundaries for each @xmath128 and then picking the result with smallest distortion , width=288 ] figure [ fig : boundaries ] presents an example of quantization boundaries .
for the given bit rate @xmath129 over the components of the input vector @xmath31 , we can express the rate over the measurements @xmath28 as @xmath130 , where @xmath131 is the measurement ratio . to determine the optimal quantizer for the given rate @xmath129 we perform optimization for all @xmath128s and return the quantizer with the least mse .
as we can see , in comparison with the uniform quantizer obtained by merely minimizing the distortion between the quantizer input and output , the one obtained via se minimization is very different ; in fact , it looks more concentrated around zero .
this is due to the fact that by minimizing se we are in fact searching for quantizers that asymptotically minimize the mse of the relaxed bp reconstruction by taking into consideration the nonlinear effects due to the method .
the trend of having more quantizer points near zero is opposite to the trend shown in @xcite for quantizers optimized for lasso reconstruction .
figure [ fig : expdist ] presents a comparison of reconstruction distortions for our two quantizers and confirms the advantage of using quantizers optimized via ( [ equ : se : serecursion ] ) . to obtain the results we vary the quantization rate from @xmath132 to @xmath133 bits per component of @xmath31 , and for each quantization rate , we optimize quantizers using the methods discussed above . for comparison
, the figure also plots the mse performance for two other reconstruction methods : linear mmse estimation and the widely - used lasso method @xcite , both assuming a bounded uniform quantizer .
the lasso performance was predicted by state evolution equations in @xcite , with the thresholding parameter optimized by the iterative approach in @xcite .
it can be seen that the proposed relaxed bp algorithm offers dramatically better performance more that @xmath134 db improvement at low rates . at higher rates ,
the gap is slightly smaller since relaxed bp performance saturates due to the awgn at the quantizer input .
similarly we can see that the mse of the quantizer optimized for the relaxed bp reconstruction is much smaller than the mse of the standard one , with more than 4 db difference for many rates .
we present relaxed belief propagation as an efficient algorithm for compressive sensing reconstruction from the quantized measurements .
we integrate ideas from recent generalization of the algorithm for arbitrary measurement channels to design a method for determining optimal quantizers under relaxed bp reconstruction .
although computationally simpler , experimental results show that under quantized measurements relaxed bp offers significantly improved performance over traditional reconstruction schemes . additionally , performance of the algorithm is further improved by using the state evolution framework to optimize the quantizers .
e. j. cands , j. romberg , and t. tao , `` robust uncertainty principles : exact signal reconstruction from highly incomplete frequency information , '' _ ieee trans .
inform . theory _
52 , pp . 489509 , feb . | we consider the optimal quantization of compressive sensing measurements following the work on generalization of relaxed belief propagation ( bp ) for arbitrary measurement channels .
relaxed bp is an iterative reconstruction scheme inspired by message passing algorithms on bipartite graphs . its asymptotic error performance can be accurately predicted and tracked through the state evolution formalism .
we utilize these results to design mean - square optimal scalar quantizers for relaxed bp signal reconstruction and empirically demonstrate the superior error performance of the resulting quantizers . |
age effects often mask metallicity effects in the studies of stellar populations @xcite , and separating them is a cumbersome affair @xcite .
the age - metallicity degeneration originates from the fact that increasing either the metallicity or the age makes the integrated spectral energy distribution ( ised ) of a single stellar population ( ssp ) redder @xcite .
previous studies showed that it is difficult to break this degeneration only by broad - band colours @xcite . in order to solve this question , some studies used spectral information , instead of colours only , in the evolutionary population synthesis ( eps ) models @xcite . with the development of these models including spectral information ,
the spectral resolution has been improved from about 20 to 2 @xcite .
furthermore , these spectral information has been translated to the line strengths either by empirical fitting polynomials @xcite or approaches other than the lick fitting functions @xcite .
the inclusion of absorption feature indices in eps models could add the power of diagnostics to the study of stellar populations @xcite , therefore some studies have included spectral absorption feature indices into eps models attempting to understand the stellar populations .
some earlier eps studies contained spectral indices include no metallicity dependence @xcite or are otherwise limited in scope @xcite .
recently , the spectral absorption indices have been combined systematically in eps studies , in those models some model builders use the absorption feature indices at intermediate resolution ( 9 ) of lick system @xcite , which has been accepted by many investigators and widely used in their studies ; other model builders use the indices at considerably higher resolution ( fwhm @xmath2 2 ) in rose system @xcite .
ssps are assemblies of chemically homogeneous and coeval single stars , the star formation history of any stellar system can be described by a superposition of ssp models with different ages and metallicities .
studying ssps can help us to understand the evolution of clusters and galaxies , the distribution of metallicities , and to quantify the star formation history of the galaxies . in this paper
we also investigate ssps with a systematic and self - consistent set of stellar evolution models by @xcite , spectral library the basel-2.0 model @xcite and the spectral absorption feature indices of lick system @xcite .
the outline of the paper is as follows . in section 2 ,
we describe our ssps models and algorithm . in section 3 ,
we give the results and discussion . in section 4
, we apply the observed spectral indices to the merit function to determine the age and the metallicity for ssp - like assembly .
finally we present the summary and conclusions in section 5 .
in this study we use the stellar evolutionary models of @xcite obtained with the eggleton stellar evolutionary code @xcite and the empirical and semi - empirical calibrated basel-2.0 stellar spectral library of @xcite . instead of using a tabular form in the previous eps studies @xcite we choose the evolutionary models of @xcite in the convenient form of the rapid single star evolution ( sse ) package presented by @xcite .
except for a set of analytic evolution functions fitted to the @xcite model tracks , the sse package extends the tracks by including a description of the remnant phases of stellar evolution , such as the white dwarf cooling track , and supplements the models of @xcite by including a prescription for mass loss , which has been neglected by the models of @xcite .
the detailed descriptions about the stellar evolutionary models of @xcite , the sse package of @xcite and the basel-2.0 stellar spectra library of @xcite have been presented in ( * ? ? ?
* paper i ) , so we do not discuss them here .
we refer the interested reader to parts 2 and 3 of paper i for them . in this paper
the factor @xmath5 ( reimers mass - loss coefficient , @xcite ) is taken to be 1/4 @xcite . except for the iseds , we will also calculate a series of lick / ids ( image dissector scanner ) spectral absorption feature indices . the lick / ids indices are the absorption strengths in feature bandpass , and the feature regions are listed in table 1 of @xcite . in lick system 21
absorption features are included , six different molecular bands ( cn4150 , g band , mgh , mgh + mgb , and two tio bands ) and 14 different blends of atomic absorption lines .
the indices of molecular bands are expressed in magnitudes , the atomic features indices are expressed in angstroms of equivalent width ( ew ) . the lick / ids absorption - line indices are given by a series of empirical fitting functions of @xcite as a function of @xmath6 , @xmath7 , and metallicity [ fe / h ] .
the effective temperature spans a range 2100 @xmath8 11000 and the metallicity is in the range @xmath9 } \leq + 0.5 $ ] .
the indices in the lick system were extracted from the spectra of 460 stars obtained between 1972 and 1984 using the red - sensitive ids and cassegrain spectrograph on the 3 m shane telescope at lick observatory .
the spectra cover the range @xmath10 , with a resolution of @xmath2 8@xcite .
for the eps of an instantaneous burst ssp the main input model parameters are : ( i ) the initial mass function ( imf ) , which gives the relative number of stars in different evolutionary stages ; ( ii ) the lower and upper mass cut - offs @xmath11 and @xmath12 ; ( iii ) the relative age , @xmath13 , of the ssp ; and ( iv ) the metallicity @xmath14 of the stars .
we use the imf of ( * ? ? ?
* hereafter ktg ) .
the ktg imf takes the form @xmath15 where a is a normalization constant and @xmath16 is the stellar mass in solar units .
the relative number of stars in the mass range @xmath17 is @xmath18 where @xmath19 is normalized by @xmath20 taking the lower and upper mass limits , @xmath11 and @xmath12 , of the stellar mass range as @xmath21 and @xmath22 respectively , gives the normalization constants @xmath23 for the ktg imf .
once the input physics database is given , we obtain evolutionary parameters such as luminosity @xmath24 , effective temperature @xmath25 , radius @xmath26 and mass @xmath16 for each star in a ssp using the sse algorithm , transform these evolutionary parameters to stellar flux with the basel-2.0 stellar spectral model , and obtain absorption feature indices of the lick / ids system using the fitting functions of @xcite . by the following equations ( [ sp - lamda ] - [ inte - mag1 ] )
we can obtain the integrated monochromatic flux and absorption feature indices for an instantaneous ssp of a particular age and metallicity . in the following equations , a parameter identified by a capital letter on the left - hand side represents the integrated ssp , while the corresponding parameter in minuscule on the right - hand side is for stars .
the integrated monochromatic flux of a ssp is defined as @xmath27 where @xmath28 is the sed of stars with mass @xmath16 and metallicity @xmath14 at a relative age of @xmath13 .
the integrated absorption feature index of the lick / ids system is a flux - weighed one . for the @xmath29th atomic absorption line
, it is expressed in equivalent width ( @xmath30 , in ) , @xmath31 where @xmath32 is the equivalent width of the @xmath29th index of stars with mass @xmath16 and metallicity @xmath14 at a relative age of @xmath13 , and @xmath33 is the continuum flux at the midpoint of the @xmath29th feature passband ; and for the @xmath29th molecular line , the feature index is expressed in magnitude , @xmath34 where @xmath35 is the magnitude of the @xmath29th index of stars with mass @xmath16 and metallicity @xmath14 at age @xmath13 . for stars cooler than 3570k , the fitting functions of @xcite use two completely different sets of coefficients for giants and dwarfs , so we must give a criteria to distinguish them .
we adopt the description of @xcite for giants and dwarfs , i.e. , @xmath36 while for stars with gravity in the range @xmath37 , @xmath38 is used to discriminate two classes of stars , i.e. , @xmath39
in this part we present the iseds at intermediate resolution ( 10 in the ultraviolet and 20 in the visible ) and the lick / ids absorption - feature indices for ssps over a large range of age and metallicity : @xmath40gyr and @xmath1 } \leq + 0.2 $ ] . as a check on our models
we compare our synthetic iseds with those of ( * ? ? ?
* hereafter gw ) and that of ngc 6838 , and compare our synthetic lick / ids absorption - feature indices with the values of @xcite , ( * ? ? ?
* hereafter vcpb)and ( * ? ? ?
* hereafter kaf ) , and those of galactic and m31 globular clusters . in figs .
[ ised - t ] and [ ised - z ] we give the variation of the intermediate resolution ised with age and metallicity over a wide wavelength range , @xmath41 .
the flux is expressed in magnitude and is normalized to zero at 2.2@xmath42 m . figs .
[ ised - t ] and [ ised - z ] show that the effects of age and metallicity on the iseds are similar , i.e. , the iseds tend to be redder with increasing age or metallicity in the wavelength region of @xmath43 .
to test our models we compare our synthetic iseds with those of gw for the young ( @xmath44gyr ) and the old ( @xmath45gyr ) ssps with solar - metallicity in the top panels of figs .
[ ised - com02 ] and [ ised - com12 ] , respectively .
the iseds of gw are based on his work in 1994 and from his private homepage @xcite , which allows one to obtain iseds and the lick / ids absorption feature indices for arbitrary mixtures of single - burst ssps with ages in the range @xmath46gyr and metallicities in the range @xmath47 } \leq + 0.5 $ ] ( but not for metallicity @xmath48 } < -0.225 $ ] with age @xmath49gyr ) . figs .
[ ised - com02 ] and [ ised - com12 ] show significant disagreement of the ised in two wavelength regions , the larger one is in the far - ultraviolet region ( @xmath50 ) , the minor is in the visible and infra - red regions ( i.e. , @xmath51 ) . in the first wavelength region
our ised is completely different from that of gw , in the second region this work exhibits bluer continuum than that of gw . comparing the discrepancy of ised in the visual and infra - red regions for the young and the old ssps it seems that this discrepancy in iseds increases with age .
0.7 cm 0.2 cm in order to investigate what actually causes the discrepancy between our iseds and those of gw , we first discuss the differences between these two eps models .
this work differs from that of gw by the adoption of * different stellar evolutionary models .
we use the stellar evolutionary models of @xcite , whereas gw use the stellar evolutionary isochrones by vandenberg and collaborators @xcite and the revised yale isochrones @xcite . and in the gw model the post asymptotic giant branch ( agb ) stars are not included . * different stellar spectral libraries .
we use the basel-2.0 library , which is based on several original grids of model atmosphere spectra : kurucz ( 1995 , private communication ) for o to late -k stars ; @xcite and @xcite for m giants in the temperature range @xmath52k ; and @xcite for m dwarfs , whereas gw used the theoretical spectra of @xcite for stars hotter than 3750k and composite spectra for cooler stars by patching together model atmospheres ( @xcite , and references therein ) and optical observational spectra of m giants @xcite . *
different imf shapes .
we use the ktg imf @xcite , whereas gw adopt the salpeter imf with a slope of @xmath53 @xcite , i.e. , @xmath54 where c is a normalization constant . the difference between our iseds and those of gw
can possibly be attributed to either the choice of stellar evolutionary models , the spectral library , the imf and/or the inclusion of post agb in eps model . in the top panels of figs .
[ ised - com02 ] and [ ised - com12 ] we supplement the ised drawn from the salpeter imf with slope @xmath53 and the ised from the ktg imf but no thermally pulsing giant branch / proto planetary nebula / planetary nebula ( tpagb / ppn / pn , these phases are beyond `` early asymptotic giant branch [ eagb ] '' , are collectively termed `` post eagb [ peagb ] '' ) stars included for solar metallicity ssps at ages of 2gyr and 12gyr .
in the bottom panels we give the fractional contributions of different evolutionary stage to the total flux for 2gyr and 12gyr ssps of solar metallicity . in it
various abbreviations are used to assign the evolution phases .
they are as follows : `` ms '' stands for main - sequence stars , `` hg '' stands for hertzsprung gap ; `` gb '' stands for the first giant branch ; `` cheb '' stands for core helium burning ; and the ms is divided into two phases to distinguish deeply or fully convective low - mass stars ( @xmath55 ) and stars of higher mass with little or no convective envelope ( @xmath56 ) .
1.0 cm 0.2 cm first , the adoption of different spectral libraries can not cause significant discrepancy of iseds in the region @xmath57 .
the reason is that the total light in this wavelength region is dominated by peagb stars ( see the bottom panels of figs .
[ ised - com02 ] and [ ised - com12 ] ) , in an exact word , by pn stars with temperature @xmath58 . for these hotter pn stars
the gw model and our model actually use the same spectral library . furthermore , the top panels of figs .
[ ised - com02 ] and [ ised - com12 ] show that ( i ) the ised of gw in the far - ultraviolet region agree with our result with the ktg imf but no peagb stars included ; ( ii ) two iseds overlap for our ssps with the different imf shapes , i.e. , the ktg imf and the salpeter imf with a slope of @xmath53 in the region @xmath57 .
therefore the significant disagreement between our ised in the region @xmath59 and that of gw is dominated by the inclusion of pn stars in peagb stage for ssps , and the imf shape is not the main factor
. 0.7 cm 0.2 cm from the bottom panel of fig .
[ ised - com02 ] we see that the total flux in the visible and infra - red regions is dominated by ms stars with mass @xmath56 , cheb stars and cooler tpagb / ppn stars in peagb stage ( only at wavelength @xmath60 ) for young ssps ( @xmath61gyr ) .
the contribution of ms stars to the light is greater than those of cheb and the cooler tpagb / ppn stars in the region of 3.5 @xmath62 4.0 , while the cooler tpagb / ppn stars make the greatest contribution to the light in the region 4.0 @xmath63 4.2 . from the bottom panel of fig .
[ ised - com12 ] we see that the light in the visible and infra - red regions is dominated by ms stars with mass @xmath56 , cheb stars and gb stars for old ssps ( @xmath64gyr ) .
the contribution of ms stars and gb stars is greater than that of the other two evolutionary stages at shorter and longer wavelengths , respectively , and the contribution of ms stars to the light is equal to gb at wavelength @xmath65 3.8 .
the top panels of figs .
[ ised - com02 ] and [ ised - com12 ] show that ( i ) the inclusion of tpagb / ppn stars in peagb stage in eps models can produce a significant deviation of iseds in the visible and infra - red regions for the young ssps ( @xmath61gyr ) , but almost no effect for the old ssps ( @xmath64gyr ) .
the reason is that the cooler tpagb / ppn stars in peagb stage contribute a lot to the light for young ssps at the red end of this region , but almost no light contribution for old ssps ( see the bottom panels of figs .
[ ised - com02 ] and [ ised - com12 ] ) .
the relatively greater contribution of the cooler tpagb / ppn stars , for young ssps , is mainly caused by the fact that the relative number of the tpagb / ppn stars is greater than that for old ssps .
( ii ) for @xmath66gyr ssps , the iseds for ssps without peagb stars included are as much as @xmath2 1mag greater than those for ssps including peagb stars .
this reason of the blueness of iseds is that the ssps would look like younger and the iseds trend to be bluer if omitting those cooler tpagb / ppn stars in peagb stage .
( iii ) also for @xmath61gyr ssps , the discrepancy in iseds arising from the inclusion of tpagb / ppn stars is larger than that arising from the different models ( the gw model and our model ) .
( iv ) the adoption of different imf shapes almost do not affect the iseds in the visible and infra - red regions .
this can be inferred from the uniqueness of iseds between the values with the ktg imf and those with the salpeter imf with slope @xmath53 .
besides the inclusion of peagb stars , the adoption of different spectral library and evolutionary models can result in the deviation of iseds in the visible and infra - red regions . in fig .
[ synch - com ] we plot theoretical isochrones for solar - metallicity ssps at ages of 2gyr and 12gyr , as calculated by gw and this work .
it reveals three significant deviations between the two isochrones : ( i ) ms stars with mass @xmath67 ( corresponding roughly to @xmath68 of 3.6 ) in our models are lightly cooler than that in the gw models ; ( ii ) sub - giant stars in our models are significantly bluer .
the cooler ms stars in our models cause redder visible and infra - red continuum , the hotter gb stars cause bluer continuum , and the reddening of iseds caused by cooler ms stars is compensated by the hotter gb stars . in summary the discrepancy between our iseds in the visible and infra - red regions and those of gw is not caused mainly by the adoption of different imf shapes , i.e. , the ktg imf and the salpeter imf with a slope of @xmath69 , but by the adoption of different stellar evolutionary models and spectral library in eps model .
the inclusion of the cooler tpabb / ppn stars in peagb stage can produce a significant deviation of iseds for young ssps , but almost no effect for old ssps in this region .
we compare model iseds with that of ngc 6838 .
the spectral data of ngc 6838 is from @xcite and covers 4000 - 6400 , its resolution is 1.25 .
the result reveals that ngc 6838 can be fitted by a ssp with solar metallicity and an age of @xmath2 12.6gyr . in fig .
[ ised - ngc6838 ] we present the ised of ngc 6838 and the model ised for an ssp with solar metallicity and an age of @xmath2 12.6gyr , the flux is expressed in magnitude and is normalized to zero at 5500 . from fig .
[ ised - ngc6838 ] we see that the agreement between model and observed spectrum is quite well in the range @xmath70 , but the agreement is not so good for shorter region . in table 1
we present all resulting lick / ids spectral absorption feature indices for 7 metallicities from @xmath71 to @xmath72 .
all indices except for @xmath73 increase with increasing age and metallicity , and the variation of these indices with age is larger at early age . in fig .
[ asi - com - auth ] we give the evolution of the lick / ids spectral absorption feature indices for solar - metallicity ssps with the ktg imf , supplemented with the indices for solar - metallicity ssps obtained by gw ( 1994 ) , vcpb ( 1996 ) and kaf ( 1999 ) . except for @xmath74 , @xmath75 , @xmath76 , @xmath77 , @xmath78 and @xmath79 indices , which show much stronger discrepancies between the studies ,
all indices agree with those of other studies .
0.7 cm 0.2 cm the vcpb model adopts the theoretical isochrones of @xcite , empirical spectra whenever possible , the fitting functions for @xmath74 and @xmath75 indices from @xcite and the fitting functions for the other lick indices from @xcite , and two imf shapes : unimodal and bimodal .
the unimodal imf takes the from @xmath80 where @xmath42 for the solar neighborhood is equal to 1.35 and @xmath81 is a constant .
the bimodal imf is described by @xmath82 where @xmath83 is a spline . for stars with mass lower than @xmath84
the tracks are from @xcite , as used by us .
kaf used the evolutionary tracks from the padova group @xcite , for lower masses ( @xmath85 ) used the tracks of @xcite , theoretical colour calibrations from @xcite , fitting functions for stellar atmospheric indices from @xcite , and a salpeter - like imf with a slope of 1.35 @xcite .
0.7 cm 0.2 cm in order to investigate which factors cause the difference between our synthetic lick / ids absorption - line indices and those of gw , especially in the @xmath74 , @xmath75 , @xmath76 , @xmath77 , @xmath78 and @xmath79 indices , in fig .
[ asi - com - auth ] we also give the lick / ids absorption feature indices for ssps with the salpeter imf with a slope of @xmath69 and the indices for ssps with the ktg imf but no peagb stars included in eps model .
it shows that the choice of the imf and the exclusion of peagb stars give rise to only small changes in @xmath86 , @xmath77 , @xmath78 and @xmath79 indices , and the deviations introduced by the imf shape and the exclusion of peagb stars are smaller than the discrepancies between the two models .
so the disagreement in the lick / ids absorption - line indices between our work and the gw model is mainly caused by the adoption of spectral library and stellar evolutionary models .
the effects of adopting different stellar evolutionary models on the lick / ids indices are as follows : ( i ) cooler ms stars with mass @xmath56 in our models ( see fig . [ synch - com ] ) make the absorption - line indices at redder end ( @xmath77 , @xmath78 , @xmath79 ) and the blue index @xmath76 higher because these indices decrease sharply with increasing temperature , but make @xmath87 and @xmath88 lower .
( ii ) hotter gb stars in our models ( see fig .
[ synch - com ] ) make @xmath77 , @xmath78 and @xmath79 and @xmath76 significantly lower , but @xmath74 and @xmath75 higher .
( iii ) the effect of cooler ms stars on the lick / ids indices is contrary to that of hotter gb stars , and the deviations are dominated by ms stars , the reason is the lick / ids indices cover the region 4000 @xmath2 6000(i.e . , 3.6 @xmath89 3.8 ) , in which the total light is dominated by ms stars with mass @xmath67 .
kaf ( 1999 ) use the same spectral library ( i.e. , the basel-2.0 library ) and the same fitting functions for lick / ids absorption indices ( i.e. the functions of worthey et al . , 1994 ) as those in our model .
so the disagreement between our absorption - line indices and those of kaf is possibly introduced by the adoption of different evolutionary models and the imf shape . in fig .
[ asi - com - auth ] we give the indices for ssps with the salpeter - like imf with a slope of @xmath90 , as used by kaf .
it shows that the deviation arising from the imf shape is smaller than that arising from the different models .
so the difference between this work and the kaf models is mainly caused by the choice of stellar evolutionary models . in fig .
[ synch - com ] the theoretical padova isochrones used by kaf are given for solar - metallicity ssps at ages of 2gyr and 12gyr .
it shows that the tracks of ms stars with @xmath56 in our models agree with those in the kaf models .
according the analysis in the above paragraph , i.e. , the deviations are dominated by ms stars with mass @xmath56 , we can draw the following conclusion : the deviations in the lick / ids indices between our models and kaf models are smaller than those between our models and gw models , in fact , this conclusion coincide with the comparison of @xmath91 and @xmath92 in fig .
[ asi - com - auth ] .
vcpb ( 1994 ) use different stellar evolutionary models , spectral library and two different imf shapes from those in our models . in fig .
[ asi - com - auth ] we give the indices from the vcpb models for the unimodal and bimodal imf with a slope of 2.35 .
it shows that the significant discrepancies between the unimodal and bimodal models exist in @xmath93 , @xmath77 , @xmath78 and @xmath78 indices .
our results are much closer to those with the bimodal imf than those with the unimodal imf .
the differences between the results in our models and those in the vcpb models are possibly caused by the adoption of different stellar evolutionary models , spectral library and the imf shape .
in fact , the vcpb and kaf models adopt the same isochrones , i.e. , padova isochrones ( see fig . [ synch - com ] ) , so the deviations of lick / ids indices introduced by the different evolutionary models between our and vcpb models , is same as those between our models and kaf models .
.manual of observational data [ cols=">,>,>,>,>,>,>",options="header " , ] [ gc - name ] as another test , the model indices are checked against a wide variety of available observations .
we compare our absorption - line indices with those of galactic and m31 globular clusters and those of the circumnuclear region of m32 in many index - index diagrams ( 210 diagrams in total , not included in this paper ) .
the line strengths of galactic and m31 globular clusters are from @xcite and that of m32 from ( * ? ? ?
* references therein ) . in the database of @xcite 36 globular clusters ( 18 m31 and 18 galactic clusters ) are included ( see table [ gc - name ] ) . m31
globular clusters are observed with the standard aperture ( 1.4@xmath94 1.4 , hereinafter sa ) .
the galactic globular clusters are observed with a long slit of standard width ( 1.4@xmath94 16 ) that was raster - scanned on the sky to create a square aperture of size 66@xmath94 66 , this resulted in two square apertures , one centered on the cluster ( hereinafter c ) and one for the `` off '' beam ( hereinafter o ) located 35 to the east . for ngc 6624
, the l aperture is 45 by 60and the s aperture is 13 by 13 , both are centered on the cluster .
raster scans have the same spectral resolution as standard - slitwidth scans ( 1.4 ) . in this paper m31 v204
has been omitted from the database of @xcite because of larger deviation from our models .
the results reveal that our models fit the data very well in many index - index diagrams , while in some index - index diagrams the models do not match real globular clusters well , e.g. in the plots of @xmath78 vs. other indices .
the index @xmath78 is systematically smaller than observations . in fig .
[ asi - com - obs ] we only give several representative plots in which the model predictions fit the data well , i.e. , @xmath73 vs. four iron lines ( @xmath95 , @xmath96 , @xmath93 and @xmath97 ) . from fig . [ asi - com - obs ] we see that there are no significant discrepancies in spectral indices between m32 and galactic globular clusters , and the required metallicity is always lower than solar and the age is older than 10gyr for most of galactic and m31 globular clusters . meanwhile from fig .
[ asi - com - obs ] we see that the metallicity of the central region of m32 approaches to solar metallicity .
our synthetic spectral absorption indices agree better with the values obtained by other authors and fit the observational data very well in most of index - index diagrams , therefore we use these spectral absorption indices to constrain or determine the age and the metallicity for ssp - like assemblies .
the method is based on minimizing a chi - squared merit function @xmath98 between all observed indices and model predictions .
the merit function is a measure for the goodness of the fit for observational data . for each model
@xmath99 merit function is defined as : @xmath100 where @xmath101 is the number of observed spectral indices , @xmath102 is the observed value of the @xmath29th index , @xmath103 is the synthetic value of corresponding index for model @xmath99 , and @xmath104 is its observational error , @xmath105 represents its relative weight . in this work
we assign the same weight ( = 1 ) to all the spectral indices . using the above merit function @xmath106 we will study the age and the metallicity of the central region of m32 , which is an elliptical galaxy . the reason we discuss m32 is as follows .
firstly , m32 is the brightest elliptical galaxy of the local group , and has been studied extensively as a template for elliptical galaxies further away .
secondly , the elliptical galaxies can be viewed as ssp - like assemblies because they appear to contain relatively little dust extinction and gaseous interstellar medium @xcite , and in them most of the stars were formed during , or very early after the initial collapse of the galaxy @xcite .
so we can investigate it by the ssps models . in fig .
[ contour ] we give its contour map of the merit function @xmath106 as a function of metallicity @xmath14 and age @xmath13 .
it shows that the data of absorption indices can be successfully fitted with an instantaneous ssp with an age of @xmath2 6.5gyr and a solar - like metallicity , with @xmath106 of 1.00 for the central region of m32 .
the age obtained by us is greater than 4gyr obtained by @xcite from their integral field spectroscopy of the 9x12 @xmath107 circumnuclear region of m32 using 30 spectral indices and colours and by @xcite from the blue spectrum using index @xmath108 , while younger than 7gyr obtained by @xcite based on their @xmath109 index and 8.5gyr from @xcite based on their hst analysis resolving m32 into individual stars .
the resulted metallicity is in full agreement with that of @xcite and @xcite , both of them predicted solar - like metallicity , but differ from that of @xcite , who predicted a metallicity of @xmath48 } = -0.25 $ ] .
applying the derived age and metallicity of the central region of m32 from merit function to the index - index diagrams ( see fig . [ contour ] ) , it seems that the plots of @xmath3 and @xmath4 are better diagrams to estimate directly the age and the metallicity from index - index plots .
the plots of @xmath3 and @xmath4 give the metallicity of @xmath110 for the center region of m32 .
in this paper we use the eps technique to present the iseds at intermediate resolution ( 10 in the ultraviolet and 20 in the visible ) and the lick / ids spectral absorption indices for an extensive set of instantaneous burst ssps of various ages and metallicities ( @xmath111gyr , @xmath1 } \leq 0.2 $ ] ) . in our eps models we adopted the rapid sse algorithm , the empirical and semi - empirical calibrated basel-2.0 spectral library and the fitting functions for the absorption - line indices . as checks on our models ,
we compare our iseds with those of worthey and the high resolution spectrum of ngc 6838 , also compare the lick / ids absorption feature indices with the values of other eps models and those of galactic and m31 globular clusters and the central region of m32 .
we conclude that : ( a ) the disagreement between our iseds with those of worthey presents in two wavelength ranges , one in the far - ultraviolet region , the other in the visible and infra - red regions . the discrepancy of iseds in the far - ultraviolet region is mainly caused by the inclusion of hotter pn stars in our models , while the discrepancy of iseds in the visible and infra - red regions is caused by the choice of stellar evolutionary models and spectral library .
the inclusion of cooler tpagb / ppn stars can produce a significant deviation in iseds for young ssps , but almost no effect for old ssps in the visible and infra - red regions .
( b ) the integrated spectrum of ngc 6838 can be fitted by a ssp with solar metallicity and an age of @xmath2 12.6gyr .
( c ) when comparing our synthetic lick / ids absorption - line indices with the values in the literatures and observational data , we find that our models are generally in agreement with other models and match real globular clusters in most index - index diagrams .
@xmath78 is , however , systematically smaller . at last using the merit function @xmath106
, we find the that spectral indices of the central region of m32 can be fitted with a single stellar population with an age of @xmath2 6.5gyr and a solar - like metallicity with @xmath112 . applying the age and the metallicity of m32 to some index - index diagrams
, we see that @xmath3 and @xmath4 are better diagrams to estimate directly the age and metallicity from index - index plots .
we acknowledge the generous support provided by the chinese natural science foundation ( grant no . 10303006 , 19925312 & 10273020 ) , the chinese academy of sciences(kjcx2-sw - t06 ) , yunnan natural science foundation ( grant no .
2002a0020q ) and by the 973 scheme ( nkbrsf g1999075406 ) .
we are deeply indebted to dr .
lejeune for making his basel-2.0 model available to us .
we also thank dr ph .
podsiadlowski for discussions .
99 aaronson m. , cohen j. g. , mould j. , malkan m. , 1978 , apj , 223 , 824 allard f. , hauschildt p. h. , 1995 , apj , 445 , 433 arimoto n. , 1996 , aps conf .
98 , from stars to galaxies , ed . c. leitherer , u. fritze - von alvensleven , & j. huchra , 287 bertelli g. , bressan a. , chiosi c. , fagotto f. , nasi e. , 1994 , a&as , 106 , 275 bessell m. s. , brett j. m. , wood p. r. , scholz m. , 1989 , a&as , 77 , 1 bessell m. s. , brett j. m. , scholz m. , wood p. r. , 1991 , a&as , 89 , 335 bressan a. , fagotto f. , bertelli g. , chiosi c. , 1993 , a&as , 100 , 647 bressan a. , chiosi c. , fagotto f. , 1994 , apjs , 94 , 63 bressan a. , chiosi c. , tantalo r. , 1996 , a&a , 311 , 425 bruzual g. , 1983 , apj , 273 , 105 bruzual g. , 1996 , revmexa&a , 4 , 61 bruzual g. , 2003 , in perez - fournon i. , balcells m. , moreno - insertis f. , sanchez f. , eds , galaxies at high redshift .
xi canary islands winter school of astrophysics .
cambridge , cambridge univ .
press , 185 bruzual g. , charlot s. , 1993 , apj , 405 , 538 buzzoni a. , 1989 , apjs , 71 , 817 buzzoni a. , 1995 , apjs , 98 , 69 buzzoni a. , gariboldi g. , mantegazza l. , 1992 , aj , 103 , 1814 buzzoni a. , chincarini g. , molinari e. , 1993 , apj , 410 , 499 carraro g. , girardi l. , bressan a. , chiosi c. , 1996 , a&a , 305 , 849 chabrier g. , baraffe i. , 1997 , a&as , 327 , 1039 charbonnel c. , meynet g. , maeder a. , schaller g. , schaerer d. , 1993 , a&a , 101 , 415 del burgo c. , peletier r. f. , vazdekis a. , arribas s. , mediavilla e. , mnras , 2001 , 321 , 227 eggleton p. p. , 1971 ,
mnras , 151 , 351 eggleton p. p. , 1972 ,
mnras , 156 , 361 eggleton p. p. , 1973 ,
mnras , 163 , 279 fagotto f. , bressan a. , bertelli g. , chiosi c. , 1994a , a&as , 104 , 365 fagotto f. , bressan a. , bertelli g. , chiosi c. , 1994b , a&as , 105 , 29 fagotto f. , bressan a. , bertelli g. , chiosi c. , 1994c , a&as , 105 , 39 fluks m. a. , plez b. , th@xmath113 p. s. , de winter d. , westerlund b. e. , steenman h. c. , 1994 , a&as , 105 , 311 frogel j. a. , persson s. e. , cohen j. g. , 1980 , apj , 240 , 785 gunn j. e. , stryker l. l. , 1983 , apjs , 52 , 121 gorgas j. , faber s. m. , burstein d. , gonzlez j. j. , courteau s. , prosser c. , 1993 , apjs , 86 , 153 green e. w. , demarque p. , king c. r. , 1987 , the revised yale isochrones and luminosity functions ( new haven : yale university observatory ) grillmair c. j. , lauer t. r. , worthey g. , faber s. m. , freedman w. l. , madore b. f. , ajhar e. a. , baum w. a. , et al .
, 1996 , aj , 112 , 1975 han z. , podsiadlowski p. , eggleton p. p. , 1994 ,
mnras , 270 , 121 hurley j. r. , pols o. r. , tout c. a. , 2000 , mnras , 315 , 543 iben i. jr . , renzini a. , 1983 , ara&a , 21 , 271 jones l. a. , worthey g. , 1995 , apj , 446 , l31 kodama t. , arimoto n. , 1997 , a&a , 320 , 41 kroupa p. , tout c. a. , gilmore g. , 1993 , mnras , 262 , 545 ( * ktg * ) kurth o. m. , alvensleben u. f .- v . , fricke k. j. , 1999 , a&as , 138 , 19 ( * kaf * ) kurucz r. l. , 1992 , the stellar populations of galaxies , iau symp .
b. barbuy & a. renzini ( dordrech : kluwer ) , 225 lejeune t. , 1997a , ph.d .
thesis , univ .
louis pasteur , strasbourg , france & astron .
institute of basel university , basel , switzerland lejeune t. , cuisinier f. , buser r. , 1997b , a&as , 125 , 229 lejeune t. , cuisinier f. , buser r. , 1998 , a&as , 130 , 65 mould j. r. , 1978 , apj , 220 , 434 mowlavi n. , schaerer d. , meynet g. , bernasconi p. a. , charbonnel c. , maeder a. , 1998 , a&as , 128 , 471 oconnell r. w. , 1976 , apj , 206 , 370 oconnell r. w. , 1986 , in stellar populations , ed . c. a. norman , a. renzini , & m. tosi ( cambridge : cambridge univ
. press ) , 167 oconnell r. w. , 1994 , in genzel t. , harris a. i , eds , nato adv . sci .
c 445 , the neclei of normal galaxies : lessons from the galactic center , dordrecht : kluwer , p. 255
peletier r. f. , 1989 , ph.d .
thesis , univ . of groningen ,
the netherland pols o. r. , tout c. a. , eggleton p. p. , han z. , 1995 , mnras , 274 , 964 pols o. r. , schr@xmath114der k. p. , hurley j. r. , tout c. a. , eggleton p. p. , 1998 ,
mnras , 298 , 525 reimers d. , 1975 , mem .
li@xmath115ge , 6e ser . , 8 , 369 renzini a. , 1981 , in effects of mass loss on stellar evolution , ed . c. chiosi & r. stalio ( dordrecht : reidel ) , 319 renzini a. , 1986 , in stellar populations , ed . c. a. norman , a. renzini , & m. tosi ( cambridge : cambridge univ . press ) , 213 rose j. a. , 1994 , aj , 106 , 107 salpeter e. e. , 1955 , apj , 121 , 161 schaller g. , schaerer d. , meynet g. , maeder a. , 1992 , a&as , 96 , 269 tantalo r. , chiosi c. , bressan a. , fagotto f. , 1996 , a&a , 311 , 361 tinsley b. m. , 1972a , a&a , 20 , 383 tinsley b. m. , 1972b , apj , 178 , 39l tinsley b. m. , gunn j. e. , 1976 , apj , 203 , 52 trager s. c. , worthey g. , faber s. m. , burstein d. , gonzalez j. j. , 1998 , apjs , 116 , 1 tripicco m. j. , bell r. a. , 1992 , aj , 103 , 1285 vandenberg d. a. , 1985 , apjs , 58 , 711 vandenberg d. a. , 1992 , apj , 391 , 685 vandenberg d. a. , bell r. a. , 1985 , apjs , 58 , 561 vandenberg d. a. , laskarides p. g. , 1987 , apjs , 64 , 103 vazdekis a. , 1999 , apj , 513 , 224 vazdekis a. , 2001 , ap&ss , 276 , 839 vazdekis a. , arimoto n. , 1999 , apj , 525 , 144 vazdekis a. , casuso e. , peletier r. f. , beckman j. e. , 1996 , apjs , 106 , 307 ( * vcpb * ) worthey g. , 1992 , ph.d . thesis , univ . of california , santa cruz worthey g. , 1994 , apjs , 95 , 107 ( * gw * ) worthey g. , astro.wsu.edu/worthey/dial/dial_a_model.html worthey g. , ottaviani d. l. , 1997 , apjs , 111 , 377 worthey g. , faber s. m. , gonzalez j. j. , burstein d. , 1994 , apjs , 94 , 687 zhang f. , han z. , li l. , hurley j. r. , 2002 , mnras , 334 , 833 ( paper i ) rrrrrrrrrrrr @xmath116 & @xmath117 & @xmath118 & @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 & @xmath124 & @xmath125 & @xmath126 & @xmath127 + @xmath128 & @xmath129 & @xmath129 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath129 + & @xmath131 & @xmath132 & @xmath133 & @xmath134 & @xmath135 & @xmath136 & @xmath137 & @xmath138 & @xmath139 & @xmath140 + & @xmath129 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath129 & @xmath129 + + @xmath116 & @xmath117 & @xmath118 & @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 & @xmath124 & @xmath125 & @xmath126 & @xmath127 + @xmath128 & @xmath129 & @xmath129 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath129 + & @xmath131 & @xmath132 & @xmath133 & @xmath134 & @xmath135 & @xmath136 & @xmath137 & @xmath138 & @xmath139 & @xmath140 + & @xmath129 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath130 & @xmath129 & @xmath129 + + + 1.0 & -0.114 & 0.089 & -0.321 & -3.184 & -2.013 & -0.026 & -0.237 & 1.208 & 5.869 & 0.645 & 0.010 + & 0.020 & 0.447 & -0.761 & 0.384 & 0.120 & 0.089 & -0.211 & 1.163 & 0.014 & 0.001 + 2.0 & -0.307 & -0.098 & -0.170 & -1.992 & -1.267 & 0.070 & 0.139 & 0.762 & 4.652 & 0.959 & 0.004 + & 0.029 & 0.593 & -0.334 & 0.504 & 0.207 & 0.118 & -0.155 & 1.147 & 0.014 & 0.002 + 3.0 & -0.288 & -0.086 & -0.091 & -1.322 & -0.873 & 0.110 & 0.293 & 0.544 & 4.082 & 1.055 & 0.002 + & 0.031 & 0.608 & -0.173 & 0.535 & 0.225 & 0.130 & -0.141 & 1.118 & 0.013 &
0.001 + 4.0 & -0.271 & -0.080 & -0.023 & -0.790 & -0.610 & 0.134 & 0.386 & 0.374 & 3.693 & 1.088 & 0.001 + & 0.030 & 0.568 & -0.080 & 0.542 & 0.224 & 0.135 & -0.139 & 1.044 & 0.012 & 0.000 + 5.0 & -0.271 & -0.087 & 0.007 & -0.669 & -0.503 & 0.137 & 0.402 & 0.294 & 3.595 & 1.070 & 0.000 + & 0.033 & 0.738 & -0.078 & 0.534 & 0.214 & 0.129 & -0.144 & 0.995 & 0.012 & -0.001 + 6.0 & -0.267 & -0.093 & 0.037 & -0.546 & -0.420 & 0.140 & 0.415 & 0.218 & 3.518 & 1.054 & 0.000 + & 0.034 & 0.796 & -0.074 & 0.529 & 0.207 & 0.124 & -0.149 & 0.959 & 0.011 & -0.002 + 7.0 & -0.269 & -0.096 & 0.053 & -0.503 & -0.387 & 0.142 & 0.421 & 0.191 & 3.480 & 1.044 & 0.000 + & 0.035 & 0.823 & -0.070 & 0.530 & 0.205 & 0.121 & -0.152 & 0.953 & 0.011 & -0.003 + 8.0 & -0.270 & -0.099 & 0.065 & -0.495 & -0.357 & 0.144 & 0.419 & 0.180 & 3.463 & 1.031 & 0.001 + & 0.036 & 0.841 & -0.067 & 0.533 & 0.205 & 0.120 & -0.155 & 0.953 & 0.011 & -0.003 + 9.0 & -0.273 & -0.102 & 0.074 & -0.509 & -0.352 & 0.142 & 0.413 & 0.171 & 3.457 & 1.016 & 0.002 + & 0.036 & 0.848 & -0.068 & 0.533 & 0.204 & 0.118 & -0.158 & 0.955 & 0.011 & -0.003 + 10.0 & -0.273 & -0.104 & 0.084 & -0.526 & -0.347 & 0.141 & 0.405 & 0.165 & 3.444 & 1.000 & 0.003 + & 0.037 & 0.857 & -0.067 & 0.535 & 0.204 & 0.116 & -0.161 & 0.963 & 0.011 & -0.004 + 11.0 & -0.274 & -0.105 & 0.094 & -0.531 & -0.321 & 0.143 & 0.402 & 0.167 & 3.416 & 0.990 & 0.004 + & 0.037 & 0.869 & -0.058 & 0.542 & 0.208 & 0.116 & -0.162 & 0.974 & 0.010 & -0.004 + 12.0 & -0.275 & -0.107 & 0.105 & -0.513 & -0.293 & 0.145 & 0.403 & 0.163 & 3.365 & 0.983 & 0.005 + & 0.038 & 0.872 & -0.046 & 0.548 & 0.211 & 0.117 & -0.162 & 0.985 & 0.010 & -0.004 + 13.0 & -0.274 & -0.108 & 0.115 & -0.523 & -0.284 & 0.145 & 0.397 & 0.166 & 3.346 & 0.970 & 0.006 + & 0.038 & 0.869 & -0.042 & 0.552 & 0.213 & 0.116 & -0.165 & 1.000 & 0.010 & -0.005 + 14.0 & -0.275 & -0.109 & 0.121 & -0.530 & -0.295 & 0.144 & 0.388 & 0.170 & 3.323 & 0.956 & 0.007 + & 0.040 & 0.894 & -0.041 & 0.555 & 0.213 & 0.115 & -0.168 & 1.013 & 0.010 & -0.005 + 15.0 & -0.277 & -0.111 & 0.131 & -0.511 & -0.261 & 0.147 & 0.384 & 0.183 & 3.287 & 0.947 & 0.009 + & 0.041 & 0.900 & -0.033 & 0.563 & 0.218 & 0.116 & -0.167 & 1.024 & 0.010 & -0.005 + 16.0 & -0.276 & -0.110 & 0.142 & -0.464 & -0.227 & 0.149 & 0.380 & 0.187 & 3.231 & 0.935 & 0.010 + & 0.041 & 0.903 & -0.028 & 0.568 & 0.220 & 0.116 & -0.169 & 1.038 & 0.010 & -0.005 + 17.0 & -0.273 & -0.109 & 0.159 & -0.358 & -0.160 & 0.154 & 0.386 & 0.178 & 3.130 & 0.930 & 0.011 + & 0.042 & 0.907 & -0.012 & 0.576 & 0.223 & 0.116 & -0.170 & 1.053 & 0.010 & -0.005 + 18.0 & -0.269 & -0.106 & 0.179 & -0.219 & -0.107 & 0.160 & 0.396 & 0.160 & 3.010 & 0.926 & 0.012 + & 0.044 & 0.944 & 0.005 & 0.584 & 0.226 & 0.115 & -0.172 & 1.070 & 0.010 & -0.005 + 19.0 & -0.151 & 0.005 & 0.199 & -0.063 & -0.026 & 0.168 & 0.409 & 0.146 & 2.878 & 0.925 & 0.013 + & 0.046 & 0.983 & 0.027 & 0.596 & 0.232 & 0.116 & -0.172 & 1.084 & 0.010 & -0.005 + + + 1.0 & -0.037 & 0.153 & -0.151 & -2.139 & -1.376 & -0.034 & 0.180 & 0.035 & 5.395 & 0.823 & 0.008 + & 0.033 & 0.645 & -0.252 & 0.324 & 0.122 & 0.132 & -0.061 & 1.109 & 0.019 & 0.012 + 2.0 & -0.219 & -0.032 & -0.014 & -1.058 & -0.891 & 0.114 & 0.602 & -0.058 & 4.494 & 1.292 & 0.006 + & 0.044 & 0.706 & 0.173 & 0.500 & 0.225 & 0.188 & 0.008 & 1.130 & 0.021 & 0.017 + 3.0 & -0.204 & -0.020 & 0.069 & -0.315 & -0.471 & 0.190 & 0.798 & -0.112 & 3.948 & 1.478 & 0.005 + & 0.047 & 0.821 & 0.354 & 0.572 & 0.266 & 0.221 & 0.033 & 1.119 & 0.020 & 0.015 + 4.0 & -0.199 & -0.018 & 0.134 & 0.269 & -0.136 & 0.247 & 0.938 & -0.153 & 3.535 & 1.600 & 0.005 + & 0.048 & 0.882 & 0.480 & 0.618 & 0.291 & 0.246 & 0.049 & 1.098 & 0.018 & 0.012 + 5.0 & -0.204 & -0.024 & 0.185 & 0.737 & 0.124 & 0.284 & 1.025 & -0.207 & 3.220 & 1.656 & 0.004 + & 0.048 & 0.950 & 0.551 & 0.639 & 0.297 & 0.260 & 0.054 & 1.045 & 0.017 & 0.010 + 6.0 & -0.210 & -0.032 & 0.232 & 1.171 & 0.378 & 0.316 & 1.098 & -0.262 & 2.911 & 1.697 & 0.003 + & 0.049 & 0.993 & 0.607 & 0.657 & 0.300 & 0.267 & 0.057 & 1.004 & 0.016 & 0.008 + 7.0 & -0.214 & -0.038 & 0.250 & 1.314 & 0.478 & 0.321 & 1.107 & -0.297 & 2.770 & 1.683 & 0.003 + & 0.050 & 1.005 & 0.615 & 0.659 & 0.296 & 0.265 & 0.055 & 0.979 & 0.015 & 0.006 + 8.0 & -0.219 & -0.044 & 0.266 & 1.393 & 0.546 & 0.325 & 1.114 & -0.319 & 2.679 & 1.673 & 0.003 + & 0.051 & 1.009 & 0.618 & 0.661 & 0.294 & 0.264 & 0.052 & 0.967 & 0.014 & 0.005 + 9.0 & -0.222 & -0.047 & 0.279 & 1.419 & 0.582 & 0.327 & 1.116 & -0.337 & 2.624 & 1.662 & 0.004 + & 0.052 & 1.002 & 0.617 & 0.661 & 0.292 & 0.262 & 0.049 & 0.966 & 0.014 & 0.004 + 10.0 & -0.224 & -0.050 & 0.287 & 1.397 & 0.721 & 0.326 & 1.112 & -0.353 & 2.584 & 1.648 & 0.004 + & 0.053 & 1.119 & 0.612 & 0.659 & 0.289 & 0.260 & 0.045 & 0.964 & 0.013 & 0.003 + 11.0 & -0.229 & -0.054 & 0.293 & 1.336 & 0.726 & 0.323 & 1.102 & -0.362 & 2.622 & 1.629 & 0.005 + & 0.053 & 1.122 & 0.605 & 0.658 & 0.287 & 0.258 & 0.042 & 0.967 & 0.013 & 0.002 + 12.0 & -0.232 & -0.057 & 0.295 & 1.243 & 0.689 & 0.316 & 1.085 & -0.372 & 2.682 & 1.606 & 0.006 + & 0.052 & 1.109 & 0.594 & 0.653 & 0.284 & 0.256 & 0.037 & 0.972 & 0.013 & 0.002 + 13.0 & -0.237 & -0.061 & 0.293 & 1.121 & 0.642 & 0.309 & 1.064 & -0.367 & 2.759 & 1.584 & 0.006 + & 0.052 & 1.100 & 0.585 & 0.653 & 0.284 & 0.255 & 0.036 & 0.978 & 0.013 & 0.001 + 14.0 & -0.241 & -0.064 & 0.292 & 1.004 & 0.582 & 0.301 & 1.041 & -0.367 & 2.822 & 1.561 & 0.008 + & 0.052 & 1.084 & 0.576 & 0.651 & 0.283 & 0.254 & 0.032 & 0.986 & 0.012 & 0.001 + 15.0 & -0.246 & -0.069 & 0.285 & 0.874 & 0.537 & 0.294 & 1.019 & -0.351 & 2.880 & 1.541 & 0.009 + & 0.052 & 1.070 & 0.570 & 0.653 & 0.285 & 0.254 & 0.032 & 0.994 & 0.012 & 0.000 + 16.0 & -0.250 & -0.072 & 0.282 & 0.765 & 0.487 & 0.287 & 0.995 & -0.340 & 2.924 & 1.518 & 0.010 + & 0.052 & 1.058 & 0.561 & 0.651 & 0.284 & 0.254 & 0.028 & 1.005 & 0.012 & 0.000 + 17.0 & -0.255 & -0.075 & 0.284 & 0.712 & 0.474 & 0.285 & 0.979 & -0.325 & 2.935 & 1.501 & 0.011 + & 0.051 & 1.025 & 0.560 & 0.654 & 0.286 & 0.255 & 0.026 & 1.017 & 0.012 & -0.001 + 18.0 & -0.257 & -0.076 & 0.289 & 0.716 & 0.476 & 0.288 & 0.973 & -0.300 & 2.914 & 1.494 & 0.012 + & 0.053 & 1.052 & 0.567 & 0.663 & 0.292 & 0.256 & 0.027 & 1.027 & 0.012 & -0.001 + 19.0 & -0.258 & -0.074 & 0.303 & 0.799 & 0.510 & 0.294 & 0.976 & -0.289 & 2.838 & 1.492 & 0.014 + & 0.055 & 1.085 & 0.579 & 0.672 & 0.296 & 0.258 & 0.026 & 1.042 & 0.012 & -0.001 + + + 1.0 & -0.140 & -0.005 & 0.029 & -1.217 & -0.826 & 0.076 & 0.655 & -0.275 & 5.046 & 1.357 & 0.016 + & 0.052 & 0.900 & 0.399 & 0.512 & 0.258 & 0.244 & 0.121 & 1.249 & 0.029 & 0.032 + 2.0 & -0.132 & -0.010 & 0.174 & -0.460 & -0.552 & 0.218 & 0.995 & -0.154 & 4.447 & 1.945 & 0.014 + & 0.059 & 1.191 & 0.670 & 0.661 & 0.338 & 0.285 & 0.140 & 1.150 & 0.036 & 0.044 + 3.0 & -0.114 & -0.004 & 0.318 & 0.797 & 0.145 & 0.371 & 1.339 & -0.092 & 3.679 & 2.245 & 0.013 + & 0.067 & 1.325 & 0.909 & 0.794 & 0.413 & 0.332 & 0.179 & 1.148 & 0.033 & 0.040 + 4.0 & -0.098 & 0.006 & 0.395 & 1.500 & 0.615 & 0.453 & 1.513 & -0.023 & 3.214 & 2.325 & 0.014 + & 0.072 & 1.386 & 1.033 & 0.885 & 0.470 & 0.359 & 0.208 & 1.193 & 0.032 & 0.037 + 5.0 & -0.087 & 0.012 & 0.456 & 2.081 & 1.019 & 0.525 & 1.650 & 0.074 & 2.841 & 2.457 & 0.016 + & 0.077 & 1.474 & 1.147 & 0.965 & 0.523 & 0.383 & 0.232 & 1.238 & 0.031 & 0.036 + 6.0 & -0.079 & 0.018 & 0.503 & 2.447 & 1.319 & 0.573 & 1.736 & 0.145 & 2.601 & 2.524 & 0.017 + & 0.082 & 1.545 & 1.223 & 1.030 & 0.567 & 0.397 & 0.250 & 1.290 & 0.030 & 0.034 + 7.0 & -0.075 & 0.020 & 0.531 & 2.722 & 1.487 & 0.600 & 1.786 & 0.151 & 2.423 & 2.566 & 0.018 + & 0.084 & 1.563 & 1.265 & 1.057 & 0.590 & 0.409 & 0.257 & 1.306 & 0.028 & 0.032 + 8.0 & -0.078 & 0.016 & 0.533 & 2.857 & 1.563 & 0.607 & 1.797 & 0.113 & 2.326 & 2.556 & 0.018 + & 0.084 & 1.556 & 1.275 & 1.053 & 0.589 & 0.417 & 0.257 & 1.289 & 0.026 & 0.027 + 9.0 & -0.084 & 0.010 & 0.526 & 2.903 & 1.585 & 0.607 & 1.795 & 0.076 & 2.281 & 2.534 & 0.018 + & 0.084 & 1.531 & 1.274 & 1.041 & 0.581 & 0.421 & 0.256 & 1.274 & 0.024 & 0.023 + 10.0 & -0.089 & 0.005 & 0.515 & 2.897 & 1.570 & 0.601 & 1.784 & 0.032 & 2.264 & 2.506 & 0.018 + & 0.085 & 1.510 & 1.266 & 1.024 & 0.568 & 0.423 & 0.253 & 1.259 & 0.022 & 0.020 + 11.0 & -0.097 & -0.002 & 0.501 & 2.846 & 1.531 & 0.593 & 1.769 & -0.009 & 2.270 & 2.475 & 0.018 + & 0.084 & 1.471 & 1.255 & 1.005 & 0.554 & 0.423 & 0.250 & 1.247 & 0.021 & 0.018 + 12.0 & -0.108 & -0.013 & 0.484 & 2.742 & 1.490 & 0.577 & 1.739 & -0.063 & 2.295 & 2.426 & 0.018 + & 0.083 & 1.438 & 1.234 & 0.984 & 0.537 & 0.422 & 0.244 & 1.230 & 0.019 & 0.015 + 13.0 & -0.112 & -0.017 & 0.473 & 2.636 & 1.595 & 0.569 & 1.725 & -0.087 & 2.328 & 2.404 & 0.018 + & 0.082 & 1.486 & 1.223 & 0.952 & 0.507 & 0.421 & 0.242 & 1.233 & 0.019 & 0.014 + 14.0 & -0.118 & -0.021 & 0.470 & 2.504 & 1.541 & 0.559 & 1.705 & -0.114 & 2.409 & 2.379 & 0.019 + & 0.082 & 1.496 & 1.210 & 0.944 & 0.503 & 0.420 & 0.239 & 1.235 & 0.018 & 0.012 + 15.0 & -0.123 & -0.025 & 0.466 & 2.348 & 1.473 & 0.547 & 1.681 & -0.140 & 2.501 & 2.348 & 0.020 + & 0.082 & 1.505 & 1.197 & 0.938 & 0.500 & 0.420 & 0.236 & 1.241 & 0.017 & 0.011 + 16.0 & -0.130 & -0.031 & 0.451 & 2.128 & 1.372 & 0.527 & 1.641 & -0.168 & 2.627 & 2.305 & 0.020 + & 0.083 & 1.502 & 1.174 & 0.927 & 0.494 & 0.417 & 0.234 & 1.244 & 0.017 & 0.010 + 17.0 & -0.137 & -0.036 & 0.436 & 1.909 & 1.264 & 0.507 & 1.599 & -0.197 & 2.739 & 2.264 & 0.021 + & 0.083 & 1.495 & 1.153 & 0.916 & 0.489 & 0.416 & 0.230 & 1.247 & 0.016 & 0.009 + 18.0 & -0.144 & -0.042 & 0.421 & 1.703 & 1.181 & 0.490 & 1.563 & -0.205 & 2.827 & 2.230 & 0.022 + & 0.083 & 1.494 & 1.139 & 0.910 & 0.487 & 0.416 & 0.228 & 1.254 & 0.016 & 0.008 + 19.0 & -0.152 & -0.048 & 0.414 & 1.561 & 1.123 & 0.480 & 1.535 & -0.208 & 2.881 & 2.201 & 0.022 + & 0.083 & 1.499 & 1.130 & 0.908 & 0.487 & 0.416 & 0.226 & 1.262 & 0.015 & 0.007 + + + 1.0 & -0.182 & -0.110 & 0.197 & -0.812 & -0.327 & 0.252 & 1.117 & -0.009 & 5.412 & 2.266 & 0.012 + & 0.067 & 1.324 & 0.896 & 0.695 & 0.316 & 0.319 & 0.214 & 1.283 & 0.038 & 0.046 + 2.0 & -0.126 & -0.071 & 0.370 & 0.988 & 0.521 & 0.526 & 1.644 & 0.460 & 4.130 & 3.058 & 0.014 + & 0.083 & 1.599 & 1.311 & 0.963 & 0.514 & 0.465 & 0.305 & 1.224 & 0.036 & 0.042 + 3.0 & -0.089 & -0.044 & 0.496 & 2.228 & 1.315 & 0.708 & 1.965 & 0.905 & 3.256 & 3.465 & 0.019 + & 0.096 & 1.834 & 1.575 & 1.163 & 0.652 & 0.540 & 0.366 & 1.320 & 0.034 & 0.039 + 4.0 & -0.064 & -0.026 & 0.580 & 3.017 & 1.861 & 0.826 & 2.160 & 1.198 & 2.739 & 3.711 & 0.024 + & 0.105 & 1.980 & 1.741 & 1.294 & 0.746 & 0.589 & 0.406 & 1.394 & 0.032 & 0.038 + 5.0 & -0.049 & -0.014 & 0.641 & 3.528 & 2.231 & 0.906 & 2.290 & 1.394 & 2.423 & 3.878 & 0.028 + & 0.112 & 2.087 & 1.853 & 1.383 & 0.812 & 0.625 & 0.434 & 1.451 & 0.032 & 0.037 + 6.0 & -0.038 & -0.005 & 0.687 & 3.877 & 2.486 & 0.967 & 2.384 & 1.553 & 2.229 & 4.024 & 0.031 + & 0.117 & 2.171 & 1.936 & 1.450 & 0.862 & 0.648 & 0.454 & 1.502 & 0.033 & 0.039 + 7.0 & -0.033 & -0.001 & 0.722 & 4.075 & 2.616 & 1.003 & 2.436 & 1.642 & 2.134 & 4.115 & 0.033 + & 0.121 & 2.221 & 1.982 & 1.491 & 0.889 & 0.656 & 0.462 & 1.548 & 0.033 & 0.041 + 8.0 & -0.027 & 0.004 & 0.763 & 4.285 & 2.744 & 1.044 & 2.492 & 1.751 & 2.050 & 4.241 & 0.035 + & 0.124 & 2.282 & 2.034 & 1.536 & 0.917 & 0.662 & 0.468 & 1.601 & 0.035 & 0.044 + 9.0 & -0.023 & 0.006 & 0.790 & 4.444 & 2.857 & 1.071 & 2.532 & 1.799 & 1.978 & 4.294 & 0.037 + & 0.129 & 2.358 & 2.069 & 1.567 & 0.938 & 0.671 & 0.474 & 1.639 & 0.035 & 0.045 + 10.0 & -0.021 & 0.008 & 0.809 & 4.571 & 2.944 & 1.091 & 2.562 & 1.814 & 1.917 & 4.319 & 0.039 + & 0.132 & 2.407 & 2.095 & 1.589 & 0.956 & 0.679 & 0.479 & 1.669 & 0.035 & 0.045 + 11.0 & -0.019 & 0.009 & 0.824 & 4.674 & 3.014 & 1.108 & 2.587 & 1.812 & 1.865 & 4.332 & 0.041 + & 0.135 & 2.441 & 2.116 & 1.607 & 0.970 & 0.687 & 0.482 & 1.696 & 0.035 & 0.044 + 12.0 & -0.018 & 0.009 & 0.836 & 4.762 & 3.068 & 1.121 & 2.606 & 1.791 & 1.818 & 4.333 & 0.043 + & 0.137 & 2.465 & 2.132 & 1.622 & 0.982 & 0.695 & 0.484 & 1.719 & 0.034 & 0.044 + 13.0 & -0.019 & 0.008 & 0.856 & 4.805 & 3.188 & 1.135 & 2.630 & 1.841 & 1.788 & 4.304 & 0.045 + & 0.141 & 2.539 & 2.156 & 1.657 & 1.002 & 0.696 & 0.491 & 1.784 & 0.034 & 0.044 + 14.0 & -0.019 & 0.007 & 0.872 & 4.839 & 3.291 & 1.146 & 2.650 & 1.876 & 1.764 & 4.268 & 0.047 + & 0.144 & 2.598 & 2.175 & 1.686 & 1.018 & 0.697 & 0.497 & 1.840 & 0.034 & 0.043 + 15.0 & -0.020 & 0.007 & 0.883 & 4.868 & 3.371 & 1.153 & 2.666 & 1.882 & 1.742 & 4.222 & 0.049 + & 0.147 & 2.631 & 2.188 & 1.706 & 1.031 & 0.698 & 0.502 & 1.884 & 0.033 & 0.042 + 16.0 & -0.020 & 0.006 & 0.886 & 4.888 & 3.431 & 1.155 & 2.676 & 1.856 & 1.722 & 4.157 & 0.051 + & 0.148 & 2.633 & 2.194 & 1.719 & 1.041 & 0.702 & 0.507 & 1.917 & 0.032 & 0.040 + 17.0 & -0.021 & 0.005 & 0.888 & 4.906 & 3.479 & 1.156 & 2.683 & 1.826 & 1.702 & 4.088 & 0.052 + & 0.151 & 2.640 & 2.197 & 1.729 & 1.049 & 0.704 & 0.511 & 1.947 & 0.030 & 0.037 + 18.0 & -0.022 & 0.004 & 0.887 & 4.889 & 3.526 & 1.152 & 2.686 & 1.792 & 1.702 & 4.011 & 0.054 + & 0.152 & 2.631 & 2.196 & 1.737 & 1.055 & 0.704 & 0.514 & 1.977 & 0.029 & 0.035 + 19.0 & -0.022 & 0.004 & 0.884 & 4.886 & 3.540 & 1.149 & 2.688 & 1.750 & 1.692 & 3.950 & 0.054 + & 0.154 & 2.624 & 2.192 & 1.739 & 1.058 & 0.705 & 0.517 & 1.995 & 0.028 & 0.033 + + + 1.0 & -0.150 & -0.088 & 0.305 & 0.242 & 0.491 & 0.504 & 1.575 & 0.664 & 5.007 & 2.855 & 0.014 + & 0.078 & 1.375 & 1.315 & 1.000 & 0.536 & 0.536 & 0.354 & 1.254 & 0.024 & 0.018 + 2.0 & -0.075 & -0.032 & 0.583 & 2.560 & 2.011 & 0.888 & 2.253 & 1.925 & 3.286 & 3.857 & 0.027 + & 0.115 & 2.022 & 1.887 & 1.506 & 0.880 & 0.676 & 0.508 & 1.650 & 0.029 & 0.031 + 3.0 & -0.042 & -0.005 & 0.708 & 3.605 & 2.833 & 1.063 & 2.537 & 2.467 & 2.606 & 4.212 & 0.039 + & 0.134 & 2.252 & 2.136 & 1.728 & 1.050 & 0.762 & 0.578 & 1.837 & 0.028 & 0.031 + 4.0 & -0.027 & 0.007 & 0.774 & 4.011 & 3.230 & 1.147 & 2.672 & 2.718 & 2.349 & 4.365 & 0.046 + & 0.145 & 2.394 & 2.257 & 1.839 & 1.141 & 0.806 & 0.617 & 1.947 & 0.028 & 0.032 + 5.0 & -0.015 & 0.018 & 0.829 & 4.367 & 3.606 & 1.218 & 2.785 & 2.927 & 2.177 & 4.487 & 0.051 + & 0.154 & 2.501 & 2.354 & 1.929 & 1.214 & 0.839 & 0.646 & 2.038 & 0.028 & 0.033 + 6.0 & -0.007 & 0.025 & 0.871 & 4.590 & 3.831 & 1.269 & 2.865 & 3.068 & 2.061 & 4.573 & 0.056 + & 0.162 & 2.605 & 2.424 & 1.995 & 1.268 & 0.862 & 0.668 & 2.109 & 0.028 & 0.034 + 7.0 & 0.000 & 0.031 & 0.901 & 4.730 & 3.992 & 1.310 & 2.929 & 3.185 & 1.979 & 4.647 & 0.060 + & 0.168 & 2.674 & 2.480 & 2.048 & 1.314 & 0.879 & 0.687 & 2.169 & 0.029 & 0.036 + 8.0 & 0.002 & 0.034 & 0.923 & 4.816 & 4.086 & 1.332 & 2.963 & 3.229 & 1.929 & 4.677 & 0.062 + & 0.172 & 2.719 & 2.510 & 2.077 & 1.336 & 0.887 & 0.693 & 2.213 & 0.029 & 0.036 + 9.0 & 0.007 & 0.038 & 0.950 & 4.963 & 4.225 & 1.364 & 3.011 & 3.305 & 1.857 & 4.728 & 0.065 + & 0.177 & 2.788 & 2.550 & 2.114 & 1.365 & 0.897 & 0.702 & 2.264 & 0.029 & 0.037 + 10.0 & 0.012 & 0.042 & 0.976 & 5.107 & 4.361 & 1.396 & 3.057 & 3.375 & 1.787 & 4.779 & 0.068 + & 0.182 & 2.852 & 2.590 & 2.151 & 1.393 & 0.906 & 0.710 & 2.312 & 0.030 & 0.038 + 11.0 & 0.016 & 0.045 & 1.000 & 5.227 & 4.465 & 1.422 & 3.094 & 3.427 & 1.728 & 4.820 & 0.070 + & 0.186 & 2.913 & 2.623 & 2.181 & 1.415 & 0.913 & 0.716 & 2.356 & 0.030 & 0.038 + 12.0 & 0.018 & 0.047 & 1.016 & 5.305 & 4.549 & 1.442 & 3.121 & 3.455 & 1.684 & 4.848 & 0.072 + & 0.189 & 2.950 & 2.647 & 2.204 & 1.432 & 0.918 & 0.719 & 2.396 & 0.030 & 0.039 + 13.0 & 0.020 & 0.048 & 1.033 & 5.386 & 4.622 & 1.461 & 3.149 & 3.474 & 1.641 & 4.875 & 0.074 + & 0.192 & 2.978 & 2.672 & 2.226 & 1.449 & 0.924 & 0.722 & 2.432 & 0.030 & 0.039 + 14.0 & 0.022 & 0.050 & 1.047 & 5.458 & 4.685 & 1.478 & 3.172 & 3.496 & 1.600 & 4.899 & 0.076 + & 0.195 & 3.018 & 2.691 & 2.245 & 1.463 & 0.927 & 0.724 & 2.467 & 0.031 & 0.040 + 15.0 & 0.022 & 0.050 & 1.058 & 5.506 & 4.731 & 1.492 & 3.190 & 3.498 & 1.569 & 4.918 & 0.078 + & 0.198 & 3.045 & 2.709 & 2.263 & 1.475 & 0.931 & 0.725 & 2.500 & 0.031 & 0.041 + 16.0 & 0.023 & 0.050 & 1.069 & 5.558 & 4.768 & 1.505 & 3.207 & 3.492 & 1.537 & 4.936 & 0.080 + & 0.200 & 3.065 & 2.725 & 2.277 & 1.486 & 0.935 & 0.726 & 2.529 & 0.031 & 0.041 + 17.0 & 0.023 & 0.050 & 1.076 & 5.601 & 4.792 & 1.516 & 3.221 & 3.470 & 1.506 & 4.951 & 0.082 + & 0.202 & 3.071 & 2.738 & 2.289 & 1.495 & 0.939 & 0.726 & 2.556 & 0.031 & 0.041 + 18.0 & 0.018 & 0.045 & 1.100 & 5.607 & 4.892 & 1.527 & 3.242 & 3.521 & 1.496 & 4.922 & 0.084 + & 0.205 & 3.159 & 2.754 & 2.321 & 1.507 & 0.933 & 0.725 & 2.627 & 0.031 & 0.042 + 19.0 & 0.016 & 0.043 & 1.118 & 5.625 & 4.957 & 1.537 & 3.260 & 3.547 & 1.480 & 4.905 & 0.086 + & 0.207 & 3.219 & 2.766 & 2.344 & 1.517 & 0.929 & 0.725 & 2.680 & 0.031 & 0.043 + + + 1.0 & -0.112 & -0.057 & 0.464 & 1.277 & 1.462 & 0.800 & 2.046 & 2.048 & 4.472 & 3.634 & 0.023 + & 0.104 & 1.709 & 1.774 & 1.494 & 0.848 & 0.698 & 0.520 & 1.700 & 0.022 & 0.017 + 2.0 & -0.046 & -0.007 & 0.773 & 3.418 & 3.088 & 1.170 & 2.678 & 3.326 & 2.846 & 4.483 & 0.048 + & 0.149 & 2.368 & 2.310 & 2.016 & 1.217 & 0.857 & 0.673 & 2.252 & 0.026 & 0.029 + 3.0 & -0.017 & 0.018 & 0.906 & 4.138 & 3.986 & 1.327 & 2.931 & 3.905 & 2.414 & 4.797 & 0.062 + & 0.175 & 2.698 & 2.540 & 2.236 & 1.392 & 0.933 & 0.746 & 2.513 & 0.027 & 0.034 + 4.0 & 0.000 & 0.034 & 0.989 & 4.534 & 4.427 & 1.425 & 3.086 & 4.266 & 2.189 & 4.985 & 0.071 + & 0.191 & 2.922 & 2.681 & 2.369 & 1.502 & 0.977 & 0.789 & 2.675 & 0.029 & 0.037 + 5.0 & 0.014 & 0.047 & 1.057 & 4.836 & 4.774 & 1.504 & 3.208 & 4.542 & 2.023 & 5.128 & 0.079 + & 0.204 & 3.091 & 2.788 & 2.472 & 1.585 & 1.009 & 0.822 & 2.806 & 0.030 & 0.040 + 6.0 & 0.025 & 0.058 & 1.118 & 5.084 & 5.059 & 1.569 & 3.309 & 4.767 & 1.893 & 5.244 & 0.085 + & 0.214 & 3.230 & 2.875 & 2.554 & 1.653 & 1.032 & 0.847 & 2.916 & 0.031 & 0.042 + 7.0 & 0.034 & 0.066 & 1.166 & 5.270 & 5.283 & 1.621 & 3.387 & 4.941 & 1.794 & 5.328 & 0.090 + & 0.223 & 3.337 & 2.941 & 2.618 & 1.704 & 1.050 & 0.866 & 3.003 & 0.031 & 0.045 + 8.0 & 0.042 & 0.074 & 1.206 & 5.423 & 5.473 & 1.666 & 3.455 & 5.090 & 1.712 & 5.403 & 0.094 + & 0.230 & 3.422 & 2.999 & 2.672 & 1.749 & 1.065 & 0.882 & 3.076 & 0.032 & 0.046 + 9.0 & 0.049 & 0.081 & 1.246 & 5.555 & 5.641 & 1.708 & 3.518 & 5.228 & 1.641 & 5.480 & 0.098 + & 0.237 & 3.492 & 3.052 & 2.722 & 1.787 & 1.077 & 0.897 & 3.148 & 0.033 & 0.049 + 10.0 & 0.055 & 0.087 & 1.280 & 5.666 & 5.791 & 1.743 & 3.570 & 5.335 & 1.577 & 5.532 & 0.101 + & 0.242 & 3.567 & 3.097 & 2.764 & 1.820 & 1.086 & 0.908 & 3.214 & 0.034 & 0.050 + 11.0 & 0.061 & 0.093 & 1.311 & 5.754 & 5.917 & 1.775 & 3.616 & 5.428 & 1.526 & 5.581 & 0.104 + & 0.248 & 3.628 & 3.135 & 2.799 & 1.849 & 1.094 & 0.917 & 3.274 & 0.034 & 0.051 + 12.0 & 0.066 & 0.098 & 1.341 & 5.836 & 6.033 & 1.804 & 3.659 & 5.513 & 1.477 & 5.623 & 0.107 + & 0.252 & 3.684 & 3.172 & 2.834 & 1.877 & 1.101 & 0.925 & 3.332 & 0.034 & 0.051 + 13.0 & 0.069 & 0.102 & 1.366 & 5.894 & 6.128 & 1.829 & 3.694 & 5.580 & 1.438 & 5.661 & 0.110 + & 0.256 & 3.725 & 3.202 & 2.861 & 1.898 & 1.105 & 0.931 & 3.385 & 0.035 & 0.053 + 14.0 & 0.073 & 0.105 & 1.387 & 5.944 & 6.210 & 1.851 & 3.725 & 5.632 & 1.403 & 5.688 & 0.112 + & 0.260 & 3.763 & 3.229 & 2.886 & 1.917 & 1.109 & 0.936 & 3.434 & 0.035 & 0.053 + 15.0 & 0.076 & 0.109 & 1.405 & 5.982 & 6.278 & 1.870 & 3.752 & 5.679 & 1.374 & 5.714 & 0.114 + & 0.263 & 3.797 & 3.252 & 2.907 & 1.934 & 1.112 & 0.940 & 3.480 & 0.036 & 0.054 + 16.0 & 0.078 & 0.111 & 1.420 & 6.013 & 6.337 & 1.887 & 3.776 & 5.716 & 1.348 & 5.736 & 0.115 + & 0.265 & 3.822 & 3.274 & 2.926 & 1.949 & 1.114 & 0.943 & 3.523 & 0.036 & 0.054 + 17.0 & 0.080 & 0.114 & 1.429 & 6.029 & 6.385 & 1.900 & 3.795 & 5.744 & 1.327 & 5.752 & 0.117 + & 0.268 & 3.839 & 3.291 & 2.941 & 1.961 & 1.115 & 0.945 & 3.562 & 0.036 & 0.055 + 18.0 & 0.083 & 0.116 & 1.440 & 6.056 & 6.431 & 1.914 & 3.815 & 5.769 & 1.305 & 5.766 & 0.118 + & 0.270 & 3.854 & 3.309 & 2.957 & 1.972 & 1.117 & 0.947 & 3.600 & 0.037 & 0.056 + 19.0 & 0.085 & 0.118 & 1.445 & 6.067 & 6.466 & 1.925 & 3.830 & 5.783 & 1.289 & 5.776 & 0.120 + & 0.271 & 3.864 & 3.323 & 2.971 & 1.983 & 1.117 & 0.948 & 3.635 & 0.037 & 0.056 + + + 1.0 & -0.094 & -0.044 & 0.565 & 1.841 & 1.973 & 0.975 & 2.299 & 2.910 & 4.163 & 4.076 & 0.031 + & 0.120 & 1.872 & 2.019 & 1.805 & 1.031 & 0.796 & 0.606 & 2.003 & 0.022 & 0.018 + 2.0 & -0.029 & 0.007 & 0.886 & 3.954 & 3.861 & 1.346 & 2.919 & 4.252 & 2.637 & 4.873 & 0.061 + & 0.172 & 2.599 & 2.542 & 2.329 & 1.415 & 0.963 & 0.756 & 2.599 & 0.026 & 0.030 + 3.0 & -0.004 & 0.029 & 1.045 & 4.534 & 4.689 & 1.499 & 3.170 & 4.842 & 2.285 & 5.173 & 0.079 + & 0.202 & 3.026 & 2.771 & 2.564 & 1.597 & 1.036 & 0.832 & 2.917 & 0.029 & 0.039 + 4.0 & 0.011 & 0.044 & 1.150 & 4.821 & 5.082 & 1.594 & 3.326 & 5.194 & 2.087 & 5.343 & 0.090 + & 0.221 & 3.266 & 2.909 & 2.706 & 1.709 & 1.076 & 0.879 & 3.124 & 0.031 & 0.045 + 5.0 & 0.024 & 0.057 & 1.242 & 5.078 & 5.427 & 1.677 & 3.459 & 5.497 & 1.923 & 5.484 & 0.100 + & 0.237 & 3.464 & 3.023 & 2.822 & 1.801 & 1.109 & 0.915 & 3.289 & 0.033 & 0.050 + 6.0 & 0.035 & 0.068 & 1.320 & 5.293 & 5.716 & 1.747 & 3.568 & 5.746 & 1.791 & 5.600 & 0.108 + & 0.250 & 3.624 & 3.114 & 2.914 & 1.874 & 1.132 & 0.942 & 3.423 & 0.034 & 0.054 + 7.0 & 0.045 & 0.077 & 1.388 & 5.455 & 5.953 & 1.805 & 3.659 & 5.952 & 1.690 & 5.695 & 0.115 + & 0.261 & 3.751 & 3.190 & 2.991 & 1.934 & 1.150 & 0.964 & 3.538 & 0.036 & 0.057 + 8.0 & 0.053 & 0.085 & 1.442 & 5.574 & 6.142 & 1.852 & 3.732 & 6.119 & 1.611 & 5.772 & 0.120 + & 0.269 & 3.850 & 3.252 & 3.051 & 1.982 & 1.165 & 0.983 & 3.631 & 0.037 & 0.059 + 9.0 & 0.061 & 0.094 & 1.495 & 5.664 & 6.311 & 1.897 & 3.803 & 6.286 & 1.543 & 5.848 & 0.126 + & 0.278 & 3.937 & 3.311 & 3.110 & 2.028 & 1.177 & 1.002 & 3.721 & 0.038 & 0.063 + 10.0 & 0.069 & 0.103 & 1.553 & 5.764 & 6.490 & 1.943 & 3.873 & 6.443 & 1.471 & 5.925 & 0.131 + & 0.286 & 4.029 & 3.372 & 3.166 & 2.074 & 1.192 & 1.020 & 3.813 & 0.039 & 0.065 + 11.0 & 0.076 & 0.110 & 1.602 & 5.840 & 6.638 & 1.981 & 3.932 & 6.569 & 1.412 & 5.984 & 0.135 + & 0.293 & 4.114 & 3.422 & 3.213 & 2.111 & 1.202 & 1.033 & 3.895 & 0.039 & 0.066 + 12.0 & 0.083 & 0.118 & 1.649 & 5.905 & 6.778 & 2.017 & 3.986 & 6.686 & 1.358 & 6.038 & 0.139 + & 0.300 & 4.184 & 3.468 & 3.256 & 2.145 & 1.211 & 1.046 & 3.972 & 0.040 & 0.067 + 13.0 & 0.090 & 0.125 & 1.695 & 5.964 & 6.907 & 2.051 & 4.038 & 6.795 & 1.308 & 6.088 & 0.143 + & 0.306 & 4.249 & 3.513 & 3.296 & 2.177 & 1.220 & 1.057 & 4.046 & 0.040 & 0.068 + 14.0 & 0.096 & 0.132 & 1.734 & 6.002 & 7.022 & 2.081 & 4.084 & 6.889 & 1.267 & 6.128 & 0.147 + & 0.311 & 4.306 & 3.552 & 3.334 & 2.207 & 1.227 & 1.068 & 4.115 & 0.041 & 0.069 + 15.0 & 0.101 & 0.137 & 1.771 & 6.040 & 7.118 & 2.108 & 4.124 & 6.967 & 1.229 & 6.165 & 0.150 + & 0.316 & 4.352 & 3.586 & 3.363 & 2.230 & 1.231 & 1.075 & 4.179 & 0.041 & 0.070 + 16.0 & 0.105 & 0.142 & 1.801 & 6.059 & 7.200 & 2.130 & 4.158 & 7.033 & 1.200 & 6.196 & 0.152 + & 0.320 & 4.394 & 3.616 & 3.390 & 2.251 & 1.234 & 1.082 & 4.238 & 0.042 & 0.071 + 17.0 & 0.110 & 0.148 & 1.832 & 6.083 & 7.283 & 2.153 & 4.193 & 7.101 & 1.168 & 6.222 & 0.155 + & 0.324 & 4.431 & 3.646 & 3.415 & 2.271 & 1.237 & 1.089 & 4.296 & 0.042 & 0.072 + 18.0 & 0.114 & 0.153 & 1.856 & 6.088 & 7.353 & 2.171 & 4.222 & 7.150 & 1.145 & 6.238 & 0.157 + & 0.327 & 4.456 & 3.672 & 3.437 & 2.288 & 1.240 & 1.095 & 4.351 & 0.042 & 0.072 + 19.0 & 0.119 & 0.158 & 1.877 & 6.094 & 7.419 & 2.188 & 4.250 & 7.202 & 1.123 & 6.255 & 0.159 + & 0.330 & 4.480 & 3.697 & 3.457 & 2.304 & 1.242 & 1.101 & 4.402 & 0.042 & 0.072 + | using evolutionary population synthesis ( eps ) we present integrated spectral energy distributions ( iseds ) and absorption - line indices defined by the lick observatory image dissector scanner ( referred to as lick / ids ) system , for an extensive set of instantaneous burst single stellar populations ( ssps ) .
the ages of the ssps are in the range @xmath0 and the metallicities @xmath1 } \leq + 0.2 $ ] .
our models use the rapid single stellar evolution ( sse ) algorithm of @xcite for the stellar evolutionary tracks , the empirical and semi - empirical calibrated basel-2.0 model of @xcite for the library of stellar spectra and empirical fitting functions of @xcite for the lick / ids spectral absorption feature indices . applying our synthetic lick / ids absorption - line indices to the merit function
we obtain the age and the metallicity of the central region of m32 , it can be well interpreted with an instantaneous ssp with an age of @xmath2 6.5gyr and a metallicity similar to solar . applying the derived age and the metallicity from the merit function to a number of index - index diagrams , we find that the plots of @xmath3 and @xmath4 are the best index - index diagrams from which we can directly obtain reasonable age and metallicity . [ firstpage ] stars : evolution galaxies : individual : m32 galaxies : stellar content . |
many physical systems , such as power systems and economic systems , often suffer from random changes in their parameters .
these parameter changes may result from abrupt environmental disturbances , component failures or repairs , etc .
in many cases , a markov chain provides a suitable model to describe the system parameter changes .
a markovian jump system is a hybrid system with different operation modes .
each operation mode corresponds to a deterministic system and the jumping transition from one mode to another is governed by a markov chain .
recently , markovian jump systems have received a lot of attention and many control issues have been studied , such as stability and stabilization @xcite , time delay @xcite , filtering @xcite , @xmath0 control @xcite , @xmath1 control @xcite , model reduction @xcite . for more information on markovian jump systems ,
we refer the reader to @xcite . in this paper , we consider the decentralized stabilization problem for a class of uncertain markovian jump large - scale systems .
the aim is to design a set of appropriate local feedback control laws , such that the resulting closed - loop large - scale system is stable even in the presence of uncertainties .
recently , the decentralized stabilization problem for markovian jump large - scale systems has been investigated in the literature ; see e.g. , @xcite and the references therein .
it is important to point out that the stabilizing techniques developed in @xcite and many other papers are built upon an implicit assumption that the mode information of the large - scale system must be known to all of the local controllers . in other words ,
the mode information of all the subsystems must be measured and then broadcast to every local controller .
such an assumption , however , may be unrealistic either because the broadcast of mode information among the subsystems is impossible in practice or because the implementation is expensive . to eliminate the need for broadcasting mode information , a local mode dependent control approach has been developed in @xcite . this control approach is fully decentralized .
the local controllers use only local subsystem states or outputs and local subsystem mode information to generate local control inputs . to emphasize this feature ,
this type of controller is referred to as a _ local mode dependent controller _ in @xcite .
as pointed out in @xcite , the local mode dependent control approach offers many advantages in practice .
first , it eliminates the need for broadcasting mode information among the subsystems and hence is more suitable for practical applications .
second , it significantly reduces the number of control gains and hence results in cost reduction , easier installation and maintenance . in this paper , we focus on the state feedback case of markovian jump large - scale systems and aim to build a bridge between the results in @xcite and @xcite .
we assume that each local controller is able to access and utilize mode information of its neighboring subsystems including the subsystem it controls .
this assumption is motivated by the fact that some subsystems may be close to each other in practice and hence exchange of mode information may be possible among these subsystems . under this assumption
, we develop an approach , which we call a _ neighboring mode dependent control approach _ , to stabilize markovian jump large - scale systems .
compared to the local mode dependent control approach , our approach can stabilize a wider range of large - scale systems in practice .
it is demonstrated in the numerical section that the system performance will improve as more detailed mode information is available to the local controllers .
hence the system performance achieved by our approach is better than that achieved by the local mode dependent control approach .
furthermore , both the global and the local mode dependent control approaches proposed in @xcite and @xcite can be regarded as special cases of the neighboring mode dependent control approach . _
notation : _ @xmath2 denotes the set of positive real numbers ; @xmath3 denotes the set of positive definite matrices ; @xmath4 denotes the set of real @xmath5 vectors ; @xmath6 denotes the set of real @xmath7 matrices .
$ ] denotes a block diagonal matrix with @xmath9 on the main diagonal .
@xmath10 is the identity matrix . for real symmetric matrices
@xmath11 and @xmath12 , @xmath13 ( respectively , @xmath14 ) means that @xmath15 is positive semi - definite ( respectively , positive definite ) .
@xmath16 denotes either the euclidean norm for vectors or the induced @xmath17-norm for matrices .
the superscript `` @xmath18 '' denotes transpose of a vector or a matrix .
@xmath19 denotes the expectation operator with respect to the underlying complete probability space @xmath20 .
consider a markovian jump large - scale system @xmath21 comprising @xmath22 subsystems @xmath23 , @xmath24 .
the @xmath25th subsystem @xmath23 is of the following form @xcite : @xmath26 where @xmath27 is the state , @xmath28 is the input , @xmath29 is the local uncertainty input , @xmath30 is the interconnection input , which describes the effect of the other subsystems @xmath31 , @xmath32 , on @xmath23 .
@xmath33 is the uncertainty output .
the initial state @xmath34 is denoted by @xmath35 .
the random process @xmath36 denotes the mode switching of the subsystem @xmath23 ; it takes values in a finite set @xmath37 .
the structure of @xmath23 is shown in fig .
[ system ] .
.,width=302 ] because the mode process @xmath36 describes the mode switching of the @xmath25th subsystem @xmath23 , the vector process @xmath38^{t}$ ] naturally describes the mode switching of the entire large - scale system @xmath21 .
we assume that @xmath38^{t}$ ] takes values in a set ( denoted by @xmath39 ) consisting of @xmath40 distinct vectors .
if any @xmath36 , @xmath41 , changes its value , the vector @xmath38^{t}$ ] will take a different value .
hence @xmath42 .
in addition , we have @xmath43 ( not necessarily @xmath44 , because the mode processes @xmath36 , @xmath41 , may depend on each other @xcite ) .
let @xmath45 , then a bijective mapping @xmath46 exists , because @xmath39 and @xmath47 have the same number of elements .
let @xmath48^{t})$ ] .
thus the random vector process @xmath38^{t}$ ] is transformed into the random scalar process @xmath49 , which carries the same mode information of the large - scale system @xmath21 . for this reason , @xmath49
is referred to as the global mode process in the sequel .
the inverse function @xmath50 is given by @xmath51^{t}$ ] , @xmath52 , @xmath53 , @xmath41 . then the @xmath25th element @xmath54 can be determined uniquely from the global mode @xmath55 .
therefore @xmath54 is also a function of @xmath55 and we write : @xmath56 , @xmath41 .
we assume here that @xmath49 is a stationary markov process .
the infinitesimal generator matrix of @xmath49 is @xmath57\in\mathbb{r}^{m\times m}$ ] , where @xmath58 if @xmath59 , and @xmath60 .
the initial distribution of the process @xmath49 is @xmath61^{t}$ ] with @xmath62 , @xmath63 . [ locally : squire : integrable ] given any locally square integrable signals @xmath64 , @xmath65 , @xmath66 , for any initial conditions @xmath67 , @xmath68 , the solution @xmath69 to each subsystem exists and is locally square integrable .
recall that a signal @xmath70 is said to be locally square integrable if it satisfies the condition @xmath71 for any finite time @xmath72 .
the term `` locally '' here means that square integrability is only required on bounded time intervals . the local uncertainty inputs and the interconnection inputs of the large - scale system are assumed to satisfy the following integral quadratic constraints ( iqcs ) .
[ defn : un : local ] given a set of positive definite matrices @xmath73 , @xmath41 .
a locally square integrable signal @xmath74^{t}$ ] represents an admissible local uncertainty input for the large - scale system if , given any locally square integrable signals @xmath75^{t}$ ] , @xmath76^{t}$ ] , there exists a time sequence @xmath77 , @xmath78 , such that for all @xmath79 and for all @xmath41 , @xmath80 dt\right ) \ge -x_{i0}^{t}\bar{s}_{i}x_{i0}. \label{un : local } \end{aligned}\ ] ] the set of all such admissible local uncertainty inputs is denoted by @xmath81
. [ defn : un : ic ] given a set of positive definite matrices @xmath82 , @xmath41 .
a locally square integrable signal @xmath76^{t}$ ] represents an admissible interconnection input for the large - scale system if , given any locally square integrable signals @xmath75^{t}$ ] , @xmath74^{t}$ ] , there exists a time sequence @xmath77 , @xmath78 , such that for all @xmath79 and for all @xmath41 , @xmath83 dt\right)\ge -x_{i0}^{t}\tilde{s}_{i}x_{i0}. \label{un : ic } \end{aligned}\ ] ] the set of all such admissible interconnections is denoted by @xmath84 .
we assume that the same time sequences @xmath85 are chosen in definition [ defn : un : local ] and definition [ defn : un : ic ] whenever they correspond to the same signals @xmath74^{t}$ ] , @xmath76^{t}$ ] , @xmath75^{t}$ ] .
the iqcs are used to describe relations between the input and output signals in the uncertainty blocks in fig .
[ system ] .
the constant terms on the right - hand sides of the inequalities and allow for nonzero initial conditions in the uncertainty dynamics .
these definitions can capture a broad class of uncertainties such as nonlinear , time - varying , dynamic uncertainties ; see ( * ? ? ?
* chapter 2.3 ) for details .
let @xmath86\in \mathbb{r}^{n\times n}$ ] be a given binary matrix , where @xmath87 if the mode of the subsystem @xmath31 is available to the @xmath25th local controller and @xmath88 , otherwise .
then the total mode information accessed by the @xmath25th local controller can be written as @xmath89^{t}$ ] .
a zero entry in this vector means that the mode information of the corresponding subsystem is not available .
we assume that the random vector process @xmath89^{t}$ ] takes values in a set ( denoted by @xmath90 ) consisting of @xmath91 distinct vectors .
obviously , @xmath92 , @xmath41 . let @xmath93 .
also , there exists a bijective mapping @xmath94 with @xmath95^{t})=\sigma_{i}$ ] , @xmath53 , @xmath96 , @xmath41 .
let @xmath97^{t})$ ] , @xmath41 .
it can be seen that @xmath98 contains essentially the same mode information as @xmath89^{t}$ ] .
hence @xmath98 is referred to as a neighboring mode process in the sequel .
both the global and the local mode dependent control problems studied in @xcite can be regarded as special cases of the neighboring mode dependent control problem with @xmath99 ( a matrix with all the elements being ones ) and @xmath10 , respectively . for the large - scale system with
the uncertainty constraints , , our objective is to design a neighboring mode dependent decentralized control law @xmath100 such that the resulting closed - loop large - scale system is robustly stochastically stable in the following sense .
the closed - loop large - scale system corresponding to the uncertain large - scale system , , and the controller is said to be robustly stochastically stable if there exists a finite constant @xmath101 such that @xmath102 for any @xmath103^{t}$ ] , and any uncertainties @xmath74^{t}\in\xi$ ] , @xmath76^{t}\in\pi$ ] . for convenience , a set of many - to - one mappings @xmath104 :
@xmath105 , @xmath41 , is introduced below : @xmath106\cdot\psi^{-1}(\mu ) ) .
\label{function } \end{aligned}\ ] ] note that @xmath104 , @xmath41 , are also surjective mappings .
[ exmp:1 ] suppose @xmath107 , @xmath108 .
when the mode processes @xmath36 , @xmath41 , are independent of each other , the vector set @xmath39 contains @xmath109 elements , i.e. , @xmath110^{t } : \mu_{i}=1 , 2 , i=1 , 2 , 3\}$ ] .
now we assume that the mode processes @xmath36 , @xmath41 , are subject to the constraints below : @xmath111 then @xmath39 contains only four elements , i.e. , @xmath112^{t } , [ 1 , 1 , 2]^{t } , [ 1 , 2 , 2]^{t } , [ 2 , 1 , 2]^{t}\}$ ] . thus @xmath113
. the mappings @xmath114 , @xmath115 between @xmath39 and @xmath47 can be defined as follows : @xmath116^{t } { \extendsymbol{\leftharpoondown}{-}{\rightharpoonup}{\psi^{-1}}{\psi } } 1,\quad [ 1 , 1 , 2]^{t } { \extendsymbol{\leftharpoondown}{-}{\rightharpoonup}{\psi^{-1}}{\psi } } 2,\\ & [ 1 , 2 , 2]^{t } { \extendsymbol{\leftharpoondown}{-}{\rightharpoonup}{\psi^{-1}}{\psi } } 3 , \quad [ 2 , 1 , 2]^{t } { \extendsymbol{\leftharpoondown}{-}{\rightharpoonup}{\psi^{-1}}{\psi } } 4 .
\end{matrix } \end{aligned}\ ] ] suppose , for example , that @xmath117 . then we have @xmath118^{t } , [ 1 , 2 , 0]^{t } , [ 2 , 1 , 0]^{t}\}$ ] . thus @xmath119 . the mapping @xmath120 can be defined as follows : @xmath121^{t } { \rightextendsymbol{-}{-}{\longrightarrow}{\rule{0.5cm}{0cm}}{\varphi_{1 } } } 1,\quad [ 1 , 2 , 0]^{t } { \rightextendsymbol{-}{-}{\longrightarrow}{\rule{0.5cm}{0cm}}{\varphi_{1 } } } 2 , \quad [ 2 , 1 , 0]^{t } { \rightextendsymbol{-}{-}{\longrightarrow}{\rule{0.5cm}{0cm}}{\varphi_{1 } } } 3 .
\end{matrix } \end{aligned}\ ] ] in this case , by , the many - to - one mapping @xmath122 is given by : @xmath123
in this section , we first turn to a new uncertain markovian jump large - scale system which is similar to the large - scale system .
global mode dependent stabilizing controllers are designed for this new large - scale system using the results of @xcite .
then we will show how to derive neighboring mode dependent stabilizing controllers for the large - scale system from these obtained global mode dependent controllers .
finally , all of the conditions for the existence of such neighboring mode dependent controllers are combined as a feasible lmi problem with rank constraints .
consider a new large - scale system @xmath124 comprising @xmath22 subsystems @xmath125 , @xmath41 .
the @xmath25th subsystem @xmath125 is as follows @xcite : @xmath126 \\ & \qquad + \tilde{e}_{i}(\eta(t))\tilde{\xi}_{i}(t ) + \tilde{l}_{i}(\eta(t))\tilde{r}_{i}(t ) , \\
\tilde{\zeta}_{i}(t ) & = \tilde{h}_{i}(\eta(t))\tilde{x}_{i}(t ) , \end{aligned}\right.\end{aligned}\ ] ] where @xmath127 , @xmath128 , @xmath129 , @xmath130 , @xmath131 for all @xmath52 and @xmath132 , @xmath41 .
the initial state @xmath133 , @xmath41 .
the uncertainties @xmath134 , @xmath135 , @xmath136 , @xmath41 , satisfy the following constraints , respectively .
[ new : defn : un : local ] a locally square integrable signal @xmath137^{t}$ ] represents an admissible local uncertainty input for the large - scale system if , given any locally square integrable signals @xmath138^{t}$ ] , @xmath139^{t}$ ] , @xmath140^{t}$ ] , there exists a time sequence @xmath77 , @xmath78 , such that for all @xmath79 and for all @xmath41 , @xmath141 dt\right ) \ge -\tilde{x}_{i0}^{t}\bar{s}_{i}\tilde{x}_{i0}. \label{new : un : local } \end{aligned}\ ] ] the set of all such admissible local uncertainty inputs is denoted by @xmath142 .
[ new : defn : un : ic ] a locally square integrable signal @xmath140^{t}$ ] represents an admissible interconnection input for the large - scale system if , given any locally square integrable signals @xmath138^{t}$ ] , @xmath139^{t}$ ] , @xmath137^{t}$ ] , there exists a time sequence @xmath77 , @xmath78 , such that for all @xmath79 and for all @xmath41 , @xmath143 dt\right)\ge -\tilde{x}_{i0}^{t}\tilde{s}_{i}\tilde{x}_{i0}. \label{new : un : ic } \end{aligned}\ ] ] the set of all such admissible interconnection inputs is denoted by @xmath144 .
[ dfn:1 ] suppose @xmath145 , @xmath41 , @xmath52 .
a locally square integrable signal @xmath139^{t}$ ] represents an admissible input uncertainty for the large - scale system if , for all locally square integrable signals @xmath138^{t}$ ] , @xmath137^{t}$ ] , @xmath140^{t}$ ] and for all @xmath41 , @xmath146 the set of all such admissible input uncertainties is denoted by @xmath147 .
we assume that the same sequences @xmath77 are chosen in definitions [ new : defn : un : local ] , [ new : defn : un : ic ] whenever they correspond to the same signals @xmath137^{t}$ ] , @xmath139^{t}$ ] , @xmath140^{t}$ ] , @xmath138^{t}$ ] .
furthermore , one can verify that the system has the same system matrices as the system , i.e. , @xmath148 , @xmath149 , @xmath150 , @xmath151 , @xmath152 at any time @xmath153 .
using this fact , we will show that @xmath154 , @xmath155 . for convenience ,
let @xmath156 denote the set of all locally square integrable signals of dimension @xmath157 , and let @xmath158 denote the set of all locally square integrable signals of dimension @xmath159 .
given @xmath137^{t}\in\tilde\xi$ ] . by definition
[ new : defn : un : local ] , the inequality holds for any signals @xmath138^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath139^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath140^{t}\in\mathcal{l}^{s}(t)$ ] .
this implies that the inequality holds for any @xmath138^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath140^{t}\in\mathcal{l}^{s}(t)$ ] and @xmath139^{t}\equiv 0 $ ] .
this is indeed the case defined by definition [ defn : un : local ] .
thus we have @xmath137^{t}\in\xi$ ] , i.e. , @xmath160 . to show @xmath161 , suppose @xmath74^{t}\in\xi$ ]
. then we shall prove @xmath74^{t}\in\tilde\xi$ ] . by definition [ new : defn : un : local ] , we need to prove that the inequality holds when we apply this signal @xmath74^{t}$ ] and any other signals @xmath138^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath139^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath140^{t}\in\mathcal{l}^{s}(t)$ ] to the large - scale system .
note that the two inputs @xmath138^{t}$ ] , @xmath139^{t}$ ] in the large - scale system can be considered as an equivalent input @xmath162^{t}$ ] . for any @xmath138^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath139^{t}\in\mathcal{l}^{m}(t)$ ] , we have @xmath162^{t}\in\mathcal{l}^{m}(t)$ ] .
thus , it suffices to prove that the inequality holds when we apply @xmath74^{t}$ ] and any other signals @xmath162^{t}\in\mathcal{l}^{m}(t)$ ] , @xmath140^{t}\in\mathcal{l}^{s}(t)$ ] .
but this follows directly from definition [ defn : un : local ] and the fact that @xmath74^{t}\in\xi$ ] .
hence @xmath74^{t}\in\tilde\xi$ ] and @xmath161 . therefore @xmath154 .
in a similar way , we can prove that @xmath155 .
we also mention that the values of @xmath163 , @xmath52 , @xmath41 , in definition [ dfn:1 ] can either be given appropriately in advance , or be solved from numerical computation as illustrated in theorem [ thm3 ] .
associated with the large - scale system is the quadratic cost functional as follows @xcite : @xmath164dt\right ) , \label{performance } \end{aligned}\ ] ] where @xmath165 , @xmath166 , @xmath52 , @xmath41 , are given weighting matrices .
for the large - scale system with the uncertainty constraints , , , global mode dependent stabilizing controllers can be designed using the technique developed in @xcite .
furthermore , applying these controllers to the large - scale system will yield a cost upper bound , i.e. , sup@xmath167 , @xmath168 .
this result is stated in the following theorem .
[ thm1 ] if there exist matrices @xmath169 and scalars @xmath170 , @xmath171 , @xmath172 , @xmath52 , @xmath41 , such that @xmath173 where @xmath174 and @xmath175 , then the global mode dependent controllers given by @xmath176 @xmath52 , @xmath41 , robustly stabilize the uncertain large - scale system with the uncertainty constraints , , , and achieve a bounded system cost @xmath177\tilde{x}_{i0}$ ] . after obtaining
the global mode dependent stabilizing controllers for the large - scale system , the next step is to derive neighboring mode dependent stabilizing controllers for the large - scale system .
the following result is an extension of theorem 1 in @xcite to the neighboring mode dependent control case .
the proof is similar to that of theorem 1 in @xcite and hence is omitted .
[ thm2 ] given the global mode dependent controllers which stabilize the large - scale system with the uncertainty constraints , , .
if the gains @xmath178 in the controllers are chosen to satisfy @xmath179 for all @xmath52 , @xmath180 , @xmath41 , then the neighboring mode dependent controllers stabilize the large - scale system with the uncertainty constraints , . in the following remark
, we use an example to illustrate the fact that theorem [ thm2 ] is less conservative than theorem 1 in @xcite . consider the markovian jump large - scale system in example [ exmp:1 ] .
three control gains need to be scheduled for the first local controller if using the neighboring mode dependent control approach , while two control gains are needed if using the local mode dependent control approach .
we denote the three neighboring mode dependent control gains as @xmath181 and the two local mode dependent control gains as @xmath182 . for comparison ,
given the global mode dependent control gains @xmath183 and the scalars @xmath184 , the constraints imposed on @xmath185 , @xmath186 are specified as follows based on theorem 1 in @xcite and our theorem [ thm2 ] , respectively : @xmath187 these inequalities are illustrated in fig .
[ fig1 ] where each circle denotes a euclidean ball .
@xmath183 is the center and @xmath188 the radius of the ball for @xmath189 . as shown in fig .
[ fig1 ] , the set where @xmath185 takes values is only a subset of the set where @xmath186 takes values .
hence the proposed framework provides greater flexibility in choosing control gains .
potentially , this will allow one to achieve better system performance than obtained using local mode dependent controllers .
we also mention that if the euclidean ball centered at @xmath190 does not intersect the euclidean ball centered at @xmath191 ( or @xmath192 ) , then no local mode dependent controllers exist .
however , the existence of the neighboring mode dependent controllers is not affected
. therefore our technique potentially produces less conservative results than that in @xcite .
next , the conditions in theorem [ thm1 ] and theorem [ thm2 ] are combined and recast as a rank constrained lmi problem . although rank constrained lmi problems are non - convex in general , numerical methods such as the lmirank toolbox @xcite often yield good results in solving these problems .
[ thm3 ] suppose there exist matrices @xmath169 , @xmath193 , @xmath194 and scalars @xmath195 , @xmath196 , @xmath197 , @xmath198 , @xmath199 , @xmath52 , @xmath41 , such that the following inequalities hold : @xmath200 where @xmath201 , \\
\mathcal{g}_{i13}(\mu ) & = y_{i}(\mu)\left[i\;i\;\tilde{h}_{i}^{t}(\mu)\;\cdots\;\tilde{h}_{i}^{t}(\mu)\right ] , \\
\mathcal{g}_{i22}(\mu ) & = -\operatorname{diag}[y_{i}(1),\cdots , y_{i}(\mu-1),\\ & \quad y_{i}(\mu+1),\cdots , y_{i}(m)],\\ \mathcal{g}_{i33}(\mu ) & = -\operatorname{diag}[\tilde{r}_{i}^{-1}(\mu),\bar{\beta}_{i}(\mu)i,\tilde{\tau}_{i}i,\tilde{\theta}_{1}i,\cdots,\tilde{\theta}_{i-1}i,\\ & \quad \tilde{\theta}_{i+1}i,\cdots,\tilde{\theta}_{n}i],\\ \upsilon_{i}(\mu ) & = k_{i}(\phi_{i}(\mu))+\tilde{g}_{i}^{-1}(\mu)\tilde{b}_{i}^{t}(\mu)x_{i}(\mu).\end{aligned}\ ] ] then a stabilizing controller is given by : @xmath202 , for @xmath203 , @xmath41 .
from @xmath169 , @xmath193 and , we have @xmath204 . similarly , @xmath205 . on the other hand , if is satisfied , by setting @xmath206 , @xmath207 , @xmath208 , @xmath209 , and applying the schur complement equivalence , the inequality is satisfied
. then , by theorem [ thm1 ] , the global mode dependent controllers can be designed to stabilize the large - scale system with the uncertainty constraints , , .
also , the lmi and the equation imply that @xmath210 for all @xmath52 , @xmath211 , @xmath41 .
that is , the inequality holds .
then , by theorem [ thm2 ] , the constructed controllers stabilize the large - scale system with the uncertainty constraints , . in @xcite , a control gain form has been proposed for the design of local mode dependent controllers .
that is , each local mode dependent control gain is chosen to be a weighted average of the related global mode dependent control gains .
this particular gain form is then incorporated into the coupled lmis from which the local mode dependent control gains are computed ; see theorem 3 and theorem 4 in @xcite for details . unfortunately , choosing such a gain form is not helpful in terms of an improvement in system performance , and sometimes may even result in infeasibility of the corresponding lmis .
a demonstration of this fact is given in section [ sec : ie ] .
indeed , such a gain form imposes an additional constraint and hence is not used in this paper .
consider the markovian jump large - scale system given in @xcite .
the mode information is @xmath212^{t},[1,2,2]^{t},[2,1,2]^{t},[2,2,1]^{t}\right\}}$ ] .
the initial distribution of @xmath49 is assumed to be the same as its stationary distribution @xmath213^{t}$ ] , which can be computed from the infinitesimal generator matrix @xmath214 . given a neighboring mode information pattern @xmath215 , our objective is to find the corresponding neighboring mode dependent stabilizing controllers for this large - scale system .
an upper bound on the quadratic cost is also evaluated for the resulting closed - loop large - scale system .
the main software we use is the lmirank toolbox @xcite .
the procedure is summarized as follows : 1 .
solve the optimization problem @xmath216x_{i0}<\gamma,\\ & \text{and~\eqref{eq : lmi},~\eqref{eq : theorem2lmi},~\eqref{beta_inverse},~\eqref{x_inverse}}. \end{aligned}\ ] ] if an optimal value @xmath217 is found , feasible neighboring mode dependent control gains are obtained .
2 . apply the obtained controllers to the large - scale system and compute the cost upper bound for the resulting closed - loop large - scale system .
the method for computing this upper bound is taken from @xcite .
it involves solving a worst - case performance analysis problem .
five cases are considered , i.e. , @xmath218 it can be seen that each neighboring mode information pattern contains more mode information than the preceding one .
@xmath219 corresponds to the local mode dependent control case , while @xmath220 corresponds to the global mode dependent control case . by using the preceding procedure , neighboring mode dependent
stabilizing controllers are found for each of these cases .
furthermore , if we apply the obtained controllers to the large - scale system , the cost upper bounds for the resulting closed - loop large - scale systems are shown in fig .
[ fig2 ] .
note that the cost upper bound found here in the local mode dependent control case is different from ( in fact , less than ) that in @xcite .
this is because the gain form proposed by theorem 3 in @xcite is not used in our computation .
one may also notice that the cost upper bound found in the case of @xmath221 is the same as the one in the case of @xmath220 .
we now explain why this happens . in the case of @xmath221 ,
each local controller obtains two subsystem modes directly .
in fact , the third subsystem mode can be derived from these two modes based on possible mode combinations in @xmath39 .
hence @xmath221 and @xmath220 are equivalent , in the sense that they yield the same performance .
this example demonstrates that the system achieves better ( or at least equal ) performance if more information about the subsystem modes is available to the local controllers .
it also shows that sometimes complete information about the global mode of the large - scale system may be redundant .
this paper has presented a decentralized control scheme for uncertain markovian jump large - scale systems .
the proposed controllers use local subsystem states and neighboring mode information to generate local control inputs . a computational algorithm involving rank
constrained lmis has been developed for the design of such controllers .
the developed theory is illustrated by a numerical example .
j. xiong , v. ugrinovskii , and i. r. petersen .
decentralized output feedback guaranteed cost control of uncertain markovian jump large - scale systems : local mode dependent control approach . in j.
mohammadpour and k. m. grigoriadis , editors , _ efficient modeling and control of large - scale systems _ , pages 167196 .
springer , new york , 2010 . | this paper is concerned with the decentralized stabilization problem for a class of uncertain large - scale systems with markovian jump parameters .
the controllers use local subsystem states and neighboring mode information to generate local control inputs . a sufficient condition involving rank constrained linear matrix inequalities
is proposed for the design of such controllers .
a numerical example is given to illustrate the developed theory .
, , , large - scale systems ; linear matrix inequalities ; markovian jump systems ; stabilization . |
the prevailing approach to quantum computation evolved from classical reversible algorithmic computation ( bennett 1979 , fredkin and toffoli 1982 ) , where a stored program drives a sequence of elementary logically reversible transformations .
in reversible - algorithmic computation a time - varying hamiltonian drives a sequence of unitary transformations ( benioff 1982 , feynman 1985 ) .
it was then found ( first by deutsch 1985 ) that entanglement , interference and measurement yield in principle dramatic speed - ups over the corresponding classical algorithms in solving some problems . in spite of this important result ,
this form of computation faces two possibly basic difficulties .
its speed - ups rely on quantum interference , which requires computation reversibility .
decoherence may then limit computation size below practical interest .
only two speed - ups of practical interest have been found so far ( factoring and database search ) , and none since 1996 .
reversible - algorithmic computation is not the most general form of quantum computation .
its limitations justify reconsidering quantum ground - state computation ( castagnoli 1998 , farhi et al .
2001 , kadowaki 2002 , among others ) , a formerly neglected approach still believed to be mathematically intractable .
quantum ground - state computation evolved from classical ground - state computation ( kirkpatrick & selman 1994 , among others ) , a well - developed approach competitive with algorithmic computation for solving boolean networks .
a boolean network is a set of nodes ( boolean variables ) variously connected by gates and wires that impose relations on the variables they connect ( fig .
1 ) . a boolean assignment satisfying all gates and wires is a network solution .
roughly speaking , all np problems can readily be converted to the problem of solving a boolean network . in quantum ground - state computation ,
one sets up a quantum network whose energy is minimum when all gates and wires are satisfied . in quantum annealing , one form of ground - state computation ,
coupling the network with a heat - bath of suitably decreasing temperature relaxes the network to its ground state , a mixture of solutions ( we assume with no significant restriction that there is at least one ) . measuring the node variables ( hermitian operators with eigenvalues @xmath0 and @xmath1 ) yields a solution .
it is believed that quantum annealing yields a ( still ill - defined ) speed - up over its classical counterpart .
quantum tunneling reduces the risk that the network , in its way toward the absolute energy minimum , remains trapped in local minima ( e.g. kadowaki 2002 )
. however , long simulation times seriously limit research on this approach . here
we develop a hybrid mode of computation .
we implement wires by ground - state computation .
we implement gates as algebraic identities resulting from quantum symmetries and statistics .
we show that relaxation - computation time is comparable with that an easier ( more loosely constrained ) logical network where all the gate constraints implemented by quantum symmetries are removed .
the comparison is based on a special projection method .
we show that the relaxation of the actual network can be obtained as a special projection of the relaxation of the easier comparison network .
this projection method shortcuts mathematical complexity and sheds light on the nature of this form of computation .
we conjecture that for this computation mode all hard - to - solve ( np ) networks become easy ( p ) and support this conjecture with plausible estimates .
decoherence is not expected to be as serious a problem for this computation model as for algorithmic computation since the network state is intentionally a thermal mixture during most of the computation .
this discussion of quantum computation still belongs to the realm of principles , like other literature on quantum ground - state computation , while algorithmic - reversible computation is now almost a technology .
nevertheless it is worth starting over with a new approach that might overcome fundamental limitations of algorithmic computation .
we use a network normal form composed just of wires and triodes ( fig.1 ) . each triode @xmath2 properly a partial gate connects three nodes labeled @xmath3 , @xmath4 , @xmath5 ( replaced by collective indices in fig.1 ) with the sum-2 relation @xmath6 where @xmath7 s are boolean variables and @xmath8 denotes arithmetical sum .
the three solutions are the rows of table i. each wire @xmath9 is an equality relation @xmath10 between two nodes @xmath11 ( table ii ) .
the example in fig .
1 , with @xmath12 nodes , @xmath13 wires ( lines ) , and @xmath14 triodes ( dashed triangles ) , has just one solution : @xmath15 , @xmath16 .
@xmath17{ccc } \begin{tabular}{|l|l|l| } \hline $ ] q_@xmath18q_@xmath18q_@xmath190@xmath181@xmath181@xmath191@xmath180@xmath181@xmath191@xmath181@xmath180@xmath20q_i@xmath18q_j@xmath190@xmath180@xmath191@xmath181@xmath21q_@xmath18q_@xmath18q_@xmath220@xmath180@xmath180@xmath18s@xmath190@xmath181@xmath181@xmath18t@xmath191@xmath180@xmath181@xmath18t@xmath191@xmath181@xmath180@xmath18t@xmath23 @xmath24 in the following we give idealized physical models of the various network elements .
network nodes represent qubits , which here we consider in a most general way as commuting hermitian operators with eigenvalues @xmath25 and @xmath1 . in most computation ,
the relations to be satisfied are achieved by a dynamical development . in principle
we may model any network relation by any valid physical relation , however . here
we model the triode relation by a spin identity . in the present computation model
, each node belongs to a triode @xmath26 , a spin 1 system , which might be two spins 1/2 , @xmath27 in units @xmath28 in a triplet state , with total spin vector @xmath29 , @xmath30 . for each spin 1
we define three qubits , @xmath31each representing a node of triode @xmath2 . by the composition of angular momentum ,
the three qubits of each proton pair satisfy the xor gate equation ( table iii ) .
the four rows of table iii correspond to the singlet and the three triplet states of proton pair @xmath2 , spanning the hilbert space @xmath32 .
we use @xmath33 as the network space . note that network nodes are not divided into inputs and outputs of the computation process as in algorithmic computation .
they are all simultaneously present in the network as commuting hermitian operators , related by time - independent identities that we use as gates . to simplify the physical model , we assume that the spatial wave function of each proton pair @xmath2 is frozen throughout the computation in a stable antisymmetric ground state , like that of the protons of an orthohydrogen molecule . in the triplet state
the three qubits of any triode @xmath34 obey the triode sum-2 relation of ( [ triode ] ) and table i. this follows from angular momentum composition and triplet symmetry , both extradynamical relations ; the triode hamiltonians are zero in @xmath35 .
let @xmath36 be the space spanned by the three triplet states of triode @xmath2 .
@xmath37 is the network subspace with all the triodes satisfied .
we define the frustration hamiltonian of wire @xmath38 in @xmath39 by @xmath40 where @xmath41 is a coefficient to provide the dimension of energy .
the eigenvalues of @xmath42 are zero when the wire is satisfied , @xmath41 when it is not ( the wire is then frustrated ) .
all @xmath42 commute
. therefore the network frustration hamiltonian is @xmath43where @xmath44 is the set of all wires in the network .
let @xmath45 be the exchange operator for the two protons of triode @xmath46 .
@xmath47 is symmetric under all the @xmath45 ( @xmath45 @xmath48 ) , since the @xmath49 s are .
therefore the triplet symmetry projection operator @xmath50 of the triode @xmath34 is a constant of motion of @xmath47 .
if the initial network state is in @xmath51 , under @xmath47 it remains in it .
the ground state of @xmath47 ( in @xmath52 ) hosts a mixture of network solutions , since all wires and triodes are satisfied .
@xmath53 is quadrilinear in the spin components @xmath54 . as a step toward implementation ,
we show that @xmath53 can be represented by pairwise spin - spin interactions by adjoining two idlers , auxiliary spin-@xmath55 variables @xmath56 , to the spin-1 variables @xmath57 already defined for each wire @xmath38 . for convenience
we normalize the idler variables to eigenvalues @xmath58 .
one of many suitable frustration hamiltonians for wire @xmath59 , connecting the nodes @xmath60 and @xmath61 ( using collective indices 1 , 2 ) is the bilinear form : @xmath62 .\]]this is chosen so that ( as is readily checked ) the ground - state projection operator of the wire (= nodes + idlers ) @xmath63 has energy eigenvalue 0 and includes eigenvectors @xmath64 with all five combinations of spin-1 eigenvalues that satisfy the wire , namely @xmath65 and @xmath66 , with correlated idler eigenvalues @xmath67 .
all the states where the wire is frustrated ( @xmath68 , @xmath69 ; @xmath70 , @xmath71 ) have energy @xmath72 for any values of @xmath73 .
the reduced state of the nodes in the ground state is the trace over the idlers , @xmath74 since all the wires are satisfied in the network state @xmath75 , they are satisfied in the reduced node state @xmath76 .
after the network relaxes to the ground state , we can find a solution to the network problem by simultaneously measuring both node bits @xmath77 , ignoring the idlers .
it is then straightforward to construct the network purely out of spins 1/2 with pairwise coupling , as in the ising model .
we leave the idlers alone but replace each spin 1 by the sum of two spins 1/2 with a coupling that favors the triplet ( parallel ) state over the singlet ( antiparallel ) overwhelmingly .
it is convenient to use heat - bath quanta that are distinguishable from the network quanta .
we use a photon - filled cavity with hilbert space @xmath78 .
@xmath79 is the system ( @xmath80network@xmath8bath ) space , @xmath81 is the subspace with triplet symmetry , all triodes satisfied .
we denote by @xmath82 the heat - bath hamiltonian and define the network - bath coupling in @xmath83 by @xmath84 .\]]this couples each proton spin to the small random gaussian time - varying magnetic field @xmath85 of the photon field at the site of the spin . to maintain the symmetry that we have assumed ,
we assume that the two protons of the same triode @xmath2 experience the same magnetic field and that the spatial wave functions of different proton pairs do not overlap . therefore triplet symmetry ( satisfaction of all triodes ) is a constant of motion of @xmath86 , thus also of the system hamiltonian @xmath87 ; in fact @xmath47 is already symmetric . with a suitable time - variation of @xmath85
, @xmath88 relaxes the network to its zero point .
let @xmath89 be the state of the system at time @xmath90 .
the development of @xmath89 is generated by @xmath91 according to the schrdinger equation .
the relaxation of the network state is described by the statistical operator @xmath92 , where tr@xmath93 means trace over the heat - bath degrees of freedom .
if @xmath94 starts in @xmath52 it remains in it , since both the wire frustration hamiltonian and the heat - bath coupling are symmetric .
a direct estimate of relaxation time is likely mathematically intractable , and a simulation is very long .
we take a shortcut that also sheds light on the nature of this hybrid computation .
we compare the network relaxation time with that of an easier network obtained by replacing all triodes ( table i ) by xor gates ( table iii ) : as if proton indistinguishability were suspended
each proton pair were replaced by a deuteron .
the restriction to @xmath95 vanishes : a network of xor gates and wires is loosely constrained and easy to solve .
in particular @xmath96 for all @xmath97 is always a solution .
a xor network is solvable in poly(@xmath98 ) time in classical computation and , one reasonably supposes , in the present hybrid computation also .
note that the xor gates of the comparison network are also extra - dynamical ; they represent a physical law , namely the composition of angular momentum . that the relaxation of the comparison network is quick can be plausibly seen as follows .
we first replace @xmath47 by the new hamiltonian @xmath99 .\ ] ] since each triode has exactly two nodes equal to @xmath1 ( table i ) , we have @xmath100 .
in the case of the actual network we have merely multiplied @xmath47 by a constant factor and the replacement is inessential .
the ground states of @xmath101 and @xmath102 are the same , and the respective energy landscapes are proportional .
not so for the comparison network , no longer restricted to @xmath103 . if @xmath104 , the energy landscape of @xmath105 has a gradient everywhere toward the solution @xmath106 allowed by table iii .
there are no local minima that can trap the comparison network on its way toward the absolute minimum ; thus the relaxation time of the comparison network is reasonably poly(@xmath98 ) .
we conjecture that introducing @xmath102 is unnecessary .
the xor gates , being extra - dynamical , would not affect relaxation time ; they could be removed , which would leave us with a set of independently relaxing wires , and no local minima .
the asymmetric hamiltonian of the comparison system in @xmath107 is @xmath108 , where @xmath109\]]is the asymmetric coupling .
now we have two independent random gaussian time - varying magnetic fields at each proton site , such that@xmath110 /2\]]is the actual heat - bath .
this is always possible since the sum of two gaussian distributions is also gaussian .
let @xmath111 be the state of the comparison system , whose development is generated by @xmath112 .
the comparison network relaxation is described by the statistical operator@xmath113 .
the symmetrization operator for all the proton pairs is @xmath114it projects @xmath83 on @xmath115 clearly @xmath116 , and so @xmath117 given that @xmath53is symmetric and @xmath118 is the identity in @xmath78 .
the development of the actual system ( hard triode network and bath ) in @xmath119 is driven by @xmath120 ; that of the comparison system ( easy xor network and bath ) in @xmath83 is driven by @xmath121 .
we show that , given ( 11 ) , the continuous projection on @xmath122 of the development of the comparison system yields the development of the actual system .
let @xmath89 @xmath123be an initial state of the actual system in @xmath95 , therefore @xmath124 under @xmath125 it develops into@xmath126under @xmath127 , it develops into@xmath128 in general non - symmetric .
we restore particle indistinguishability by projecting @xmath129 on @xmath95 , symmetrizing it : @xmath130we can see that the continuous projection of the comparison development yields the actual development .
computation time is by assumption poly(@xmath98 ) for the comparison easy xor network . to estimate that of the actual hard triode network , we decompose a @xmath131 into @xmath132 consecutive time slices @xmath133 @xmath134 $ ] of equal length @xmath135 within each @xmath136 we consider the relaxation of the comparison xor network in @xmath83 , described by @xmath137 . at the end of each @xmath136 we project * * * * @xmath138 * * * * on @xmath139 , then take the limit @xmath140 .
this yields the actual network relaxation @xmath141 . within each @xmath142
we consider the decomposition@xmath143 * @xmath144 describes networks with satisfied triodes and wires , namely solutions of the actual network ; its probability is @xmath145 * @xmath146 describes networks with satisfied triodes and at least one frustrated wire ; @xmath147 * @xmath148 describes networks with at least one violated triode , wires are either satisfied or frustrated ; @xmath149 we have considered all the possible states of the comparison network . therefore @xmath150 @xmath151 .
@xmath152 goes to zero with @xmath153 and is annihilated by each projection .
the actual network - bath interaction soon randomly generates a @xmath154 , a mixture of solutions of the actual network , with an extremely small probability @xmath155 @xmath156 . for a given confidence level ,
@xmath157 does not depend on @xmath98 .
for @xmath158 we apply the projection method .
@xmath159 becomes the nucleus of condensation of the network solutions . within each and every @xmath142
, we take a constant - average logarithmic rate of decrease @xmath160 of the frustration energy of the comparison network : @xmath161we will show later that there is no error in taking a constant - average rate . the relaxation time constant @xmath162 is by assumption poly(@xmath98 ) .
we have @xmath163@xmath164 .
in fact there is no contribution from @xmath165 , which is the ground state of @xmath166 and a possible contribution from@xmath167 would anyhow be second order infinitesimal .
let @xmath168 be the @xmath169-th ( population ) element of the diagonal of @xmath146 .
of course @xmath170 @xmath171 .
@xmath172 is diagonal , thus we have @xmath173 @xmath174 , where @xmath175 is the @xmath169-th diagonal element of @xmath53.therefore @xmath176 and @xmath177 go to zero together .
thus on average : @xmath178the decrease of @xmath177 implies an equal increase of @xmath179 @xmath8 @xmath180 it is reasonable and conservative to consider the increase of @xmath181 dominant .
in fact the relaxation of the comparison network is quicker because triodes can be violated .
note that we compare relaxation rates , not directions : the comparison network can head toward @xmath182 , the actual network remains in @xmath52 .
it is like comparing the speed of the keel and the wind on a broad reach .
speeds are proportional , while keel and wind go to different places . furthermore , @xmath47 does not couple @xmath144 with @xmath183 or @xmath148 .
in fact @xmath184 = @xmath185 .
therefore @xmath179 neither decreases nor increases on average .
since @xmath177 decreases and @xmath179 does not , the ratio @xmath186 decreases .
when we project on @xmath52 at the end of @xmath142 , we remain with a smaller @xmath177 and a larger @xmath179 ( probability of solutions of the actual network ) .
we can focus on the take - off of the probability of solution from the extremely small value @xmath187 to @xmath179 close to @xmath1 , say @xmath188 .
during take - off and within each @xmath142 we have @xmath189 ; @xmath190 grows from @xmath191 to @xmath192 , because of ( 17 ) and the assumption that @xmath193 remains unaltered .
the projection at the end of @xmath142 annihilates @xmath194 reducing @xmath195 by about @xmath196 renormalizing @xmath195 then multiplies @xmath179 by about @xmath197 at each @xmath142 . after a time @xmath198 and in the limit @xmath199 , we obtain for the actual network : @xmath200the probability of having solutions of the actual network becomes @xmath201 in a time @xmath202poly(@xmath98)@xmath80poly(@xmath98 ) .
using a different @xmath203 for each @xmath142 , with average value @xmath204 , yields the same result : @xmath205 in @xmath206 should be replaced by @xmath207 .
the extradynamical algebraic relations expressing particle statistics and angular momentum composition can replace the dynamical algebraic relations following from equations of motion as computational gates . in this new form of quantum computation ,
the gates of a boolean network are always satisfied as constants of the motion , leaving only equality relations ( wires ) to be implemented dynamically .
this form of quantum computation is expected to be robust , since it relies on thermal mixtures , not pure states , and is plausibly conjectured to be fast , turning all np problems in principle into p. as in quantum algorithmic computation , the speed - up is due to the extradynamical character of the computation ( castagnoli & finkelstein 2001 , 2002 ) .
this model of computation highlights the conceptual difference between how structures can be assembled in the classical and quantum domain .
quantally it is as though one could assemble a jigsaw puzzle simply by piling the pieces up and letting gravity lower them into mutual positions that solve the puzzle , analogously to quantum wire relaxation .
classically this way of assembling the pieces would be plagued by local energy minima .
one may wonder whether the assembly of biological molecules under hydrophobic pressure draws on similar quantum effects . | we develop a computation model for solving boolean networks that implements wires through quantum ground - state computation and implements gates through identities following from angular momentum algebra and statistics .
the gates are static in the sense that they contribute hamiltonian 0 and hold as constants of the motion ; only the wires are dynamic .
just as a spin 1/2 makes an ideal 1-bit memory element , a spin 1 makes an ideal 3-bit gate .
such gates cost no computation time : relaxing the wires alone solves the network .
we compare computation time with that of an easier boolean network where all the gate constraints are simply removed .
this computation model is robust with respect to decoherence and yields a generalized quantum speed - up for all np problems . |
we examine the relativistic electron populations in two wide - angle tails ( wats ) : 1231 + 674 and 1433 + 553 .
previous analyses @xcite have identified two apparently distinct populations of relativistic electrons : a flat spectrum jet and a steep spectrum sheath. the overall question for this research is : can the current aging models account for all of the observed spectra ?
we apply two techniques previously developed by katz - stone and rudnick @xcite to three frequency observations ( @xmath0@xmath0 20 , 6 , and 3.6 cm ) taken from the very large array : spectral tomography and color - color diagrams . in this paper , @xmath1 is defined as i @xmath2 @xmath3 .
spectral tomography is a technique designed to isolate overlapping structures with unique spectral indices .
two tomography galleries of images are constructed from pairs of frequency maps where ( e.g. ) : @xmath4 slices are taken on each image in the tomography gallery at the same position .
the spectral index @xmath5 of the structure ( e.g. the jet ) is determined when the jet disappears at @xmath6 = background level .
color - color diagrams are plots of spectral index at one pair of frequencies vs. spectral index at another pair of frequencies @xcite . as such
, only three frequencies are needed to plot the local curvature of the spectrum .
the locus of different power - law spectra would lie along the line where @xmath7 .
plotted on the color - color diagram in fig .
2 are the wat data for 1231 + 674 and the three standard aging models of a synchrotron spectrum .
the continuous injection ( ci ) model assumes that relativistic electrons are constantly supplied into the system .
the jaffe & perola ( jp ) model scatters the relativistic electrons in order to maintain an isotropic pitch angle distributions .
the kardashev & pacholczyk ( kp ) model maintains a constant pitch angle for the relativistic electrons .
[ .5 ] [ allcc ]
although the current aging models can not account for the observed spectra for all positions in 1231 + 674 or 1433 + 553 ( not shown here ) , both sources exhibit similar spectral characteristics : 1 . what is the injection spectrum of the jet ? + the spectral indices for the jet points ( closed circles ) in 1231 + 674 were obtained from the tomography slices .
the injection index of the flat spectrum component ( jet ) follows a power - law with a spectral index of @xmath1 @xmath8 -0.55 @xmath9 0.05 .
the fr i galaxy 3c449 @xcite also shows a similar injection spectrum of @xmath1 @xmath8 -0.53 @xmath9 0.01 .
no steepening of the jet s spectrum can be seen up to 56 kpc for 1231 + 674 and 29 kpc for 1433 + 553 .
how do the spectral shapes change with resolution ?
+ the full integrated spectrum ( open circle ) , lies close to a power - law ( figure 2 @xmath10 table 1 ) . however , this integrated spectral index @xmath11 @xmath12 ( @xmath8 -0.55 ) .
if we convolve the maps to a lower resolution than shown here in fig.1 , the observed spectra lie close to the power - law line with a broad range of power - laws .
this is due to the mixing of spectral components .
again , there is no relation to the apparent injection index ( @xmath1 @xmath8 -0.55 ) of the jet .
if we observe at higher resolution , it may be possible to determine the intrinsic shape of the spectrum for isolated components .
+ [ cols="^,^,^,^",options="header " , ] 3 . what is the relationship between the spectral shapes and the local structure of the wats ? + the squares ( sheath ) , obtained from a straight division where @xmath1 = ( ln @xmath13 / ln @xmath14 ) / ( ln @xmath15 / ln @xmath16 ) , were selected based on their low surface brightness
. this ` sheath ' population ( as defined by ) appears to represent a roughly separate spectral population in the color - color diagram . in the color - color diagram
, we find that the squares are largely consistent with a single jp spectrum with an injection index of @xmath8 -0.5 as measured for the jet .
therefore , the criterion of low surface brightness appears to be a reliable indicator of an isolated spectral population .
the triangles , also obtained from a straight division , represent every other location on the wats .
however , they are not consistent any of the standard models . they can be simply understood as an apparently broader spectrum caused by a varying mixture of flat spectrum jet and steep spectrum sheath components .
+ everywhere else in both wats , the spectral shapes are due to either a mixture of relativisitic electron populations or inhomogeneities in the local magnetic field @xcite .
a single kp or jp model is consistent with many of the spectral regions if we allow for these conditions .
in addition , an empirical model presented in for cygnus a can also represent much of the spectral shapes seen in the color - color diagrams .
for both 1231 + 674 and 1433 + 553 , the jp model ( with @xmath12 @xmath8 -0.5 ) serves as an approximate boundary for the maximum curvature seen in the spectrum .
+ alternatively , the observed spectra may be explained by using two homogeneous but distinct electron populations . in this scenario
, the jet and sheath begins with different injection indices .
both 1231 + 674 and 1433 + 553 would have @xmath17 @xmath8 -0.55 while @xmath18 @xmath8 -1.0 and -0.85 respectively .
future work will include further discussions about these population alternatives and the implications our results may have on current aging analyses . | color - color diagrams have been useful in studying the spectral shapes in radio galaxies . at the workshop we presented color - color diagrams for two wide - angle tails , 1231 + 674 and 1433 + 553 , and found that the standard aging models do not adequately represent the observed data .
although the jp and kp models can explain some of the observed points in the color - color diagram , they do not account for those found near the power - law line .
this difficulty may be attributable to several causes .
spectral tomography has been previously used to discern two separate electron populations in these sources .
the combination spectra from two such overlying components can easily resemble a range of power - laws .
in addition , any non - uniformity in the magnetic field strength can also create a power - law - like spectrum .
we will also discuss the effects that angular resolution has on the shape of the spectrum .
, , , |
colloidal dispersions are often used as model systems to investigate a variety of phenomena in ( soft ) condensed matter physics , including crystallization , clustering / gelation , and the glass transition.@xcite particles can interact with each other via e.g. excluded volume only,@xcite electrostatic interactions and van der waals attractions.@xcite however , external fields such as gravity and electric and magnetic fields can also influence the phase behavior either directly by changing the interparticle potentials , or indirectly by coupling to the osmotic pressure , or both . due to their mesoscopic sizes
, colloids have been used to study a variety of problems related to many particle statistical mechanics in- and out of equilibrium .
it is experimentally possible to use relatively simple and powerful techniques , such as optical and confocal microscopy @xcite or static and dynamic light scattering @xcite to fully characterize their structural and dynamical properties .
however , because colloids are much larger than atoms ( @xmath0 times larger ) , and the interparticle cohesion is therefore relatively weak , gravity often plays a significant role and affects both the colloidal equilibrium phase behavior , as well as dynamical processes . on the single - particle level
the gravitational length , @xmath1 , where @xmath2 is the thermal energy and @xmath3 is the buoyant colloid weight ( with @xmath4 the particle diameter , @xmath5 the density mismatch between the particle and the solvent , and @xmath6 the gravitational acceleration ) , characterizes at what heights gravity will start to play an important role in the equilibrium phase behavior .
if the gravitational length is sufficiently larger than the colloidal size the equation of state can be deduced from the density profile in a single sample . for this , the density profile should be measured as a function of height.@xcite recent work has shown that this can be obtained quite effectively from real - space microscopy techniques.@xcite however , care has to be taken in the case of systems with long - range electrostatic interactions between the colloids , as the effect of the double layer complicates this scheme.@xcite the pclet number , @xmath7 , the ratio between the effect of sedimentation against that of brownian diffusion , is another dimensionless quantity that characterizes the influence of gravity , here with its effects on the particle dynamics : @xmath8 where @xmath9 and @xmath10 are the time scales for brownian diffusion and sedimentation respectively .
when @xmath11 , gravity may be neglected , while in the case @xmath12 the behavior of a colloidal dispersion will be strongly affected by gravity . from this relation ( eq .
[ peclet ] ) , one can appreciate the strong effects of particle size and density on the pclet number as this varies with the fourth power of the particle diameter , @xmath13 , and linearly with the density mismatch , @xmath14 .
for example , for a colloidal particle of 1 @xmath15 m diameter and a density mismatch of @xmath16 g / ml ( @xmath17 m ) , the drift of the particle due to gravity is 10 times larger than that of brownian diffusion . by choosing the right particle size and matching the density of the particles to that of the solvent
, one can significantly reduce the influence of gravity in laboratory experiments .
however , it is experimentally difficult to reduce the density difference @xmath18 significantly below @xmath19 g / ml without extreme temperature control ( better than 0.01 @xmath20c ) required to reach the condition of @xmath15-gravity.@xcite in addition , the gravitational length and pclet number are single particle properties ; clearly , interactions at increased volume fractions can significantly modify the effects of gravity . for instance , when colloids form solid - like structures , such as aggregates or crystallites , the relevant buoyant force is not that of a single particle , but rather that experienced by solid structures composed of many colloids . because of the already mentioned strong size dependence of both the gravitational length and the sedimentation velocity , these collective effects are enormous and can already make the effects of gravity significant for nano - particle dispersions if they solidify
moreover , a sedimenting piece of a colloidal solid , like a crystal or a gel , experiences significant shearing forces of the liquid during its fall , and sets up a convective counter flow as well .
it was only recently realized that such flows can significantly alter the shape of the colloidal crystals formed.@xcite these effects alter the individual particle trajectories and consequently affect the growth mechanism , nucleation process and collective dynamics .
although less severe than in the case of solidification , in the case of attractive interactions in general particles will on average spend more time in a `` clustered state '' than in case of repulsive systems , thus increasing the effects of gravity .
when the attractions are large enough to cause a gas - liquid phase separation , again the newly formed denser phase will quickly start to sediment .
therefore , it is important to gauge the effects of gravity , and to determine when experiments can be considered to be in the limit of _
zero_-gravity , especially for colloidal systems with a gravitational length smaller than the container size .
experiments which compare those performed at @xmath21 with the same experiments done in a microgravity environment are ultimately required to quantify the gravity related phenomena and to investigate the influence of weak forces on physico - chemical processes , such as phase separation and aggregation / gelation .
this question is what drove scientists to perform experiments under micro - gravity conditions in so - called parabolic flights that can be performed by airplanes@xcite , and also in the space shuttle@xcite and the international space station.@xcite the main findings , with a focus on solidification and phase separation , will be briefly reviewed in the following .
although there certainly are important effects of gravity on the formation of colloidal particles as well ( see e.g. vanderhoff _ _ et al.__@xcite ) , these will not be considered here .
a few studies have been performed during the last 15 years to elucidate these effects , going from earth milligravity to space microgravity experiments .
the chaikin and russel groups performed experiments in the space shuttle columbia to study the effects of gravity on the crystallization of colloidal hard spheres.@xcite microgravity experiments performed in the coexistence region ( @xmath22 ) showed a face - centred cubic ( f.c.c . )
structure , suggesting that f.c.c .
is the equilibrium structure for colloidal hard spheres.@xcite this was confirmed by computer simulations.@xcite on the other hand , colloidal crystals grown just above the melting volume fraction ( @xmath23 ) exhibited a random stacking of hexagonally close - packed planes ( r.h.c.p . ) , whereas in normal gravity they showed a mixture of r.h.c.p . and f.c.c .
packings , suggesting that the f.c.c .
component was formed as a result of gravity - induced stresses.@xcite in the same experiments in space , dendritic growth of colloidal crystals was observed .
this kind of crystal growth was inhibited in normal gravity , because the rough dendrites were simply sheared off by the flow caused by the sedimentation of the colloidal crystals . at much higher volume fractions , glassy systems , which failed to crystallize on earth , crystallized fully in less than two weeks in microgravity .
simeonova _ _ et al.__@xcite used real - space fluorescent recovery after photobleaching ( frap ) to study the influence of gravity on the long - time behavior of the mean - square displacements ( msd ) of glassy polydisperse colloidal hard spheres . using two density matching conditions , corresponding to different gravitational lengths : @xmath24 of 0.2 mm ( low gravity effect ) and 13 @xmath15 m ( normal gravity )
, the authors showed that under the effect of normal gravity and for samples prepared at volume fractions higher than @xmath25 , aging takes place over a much wider time window .
that explains the observation that colloidal hard sphere systems which are glassy on earth rapidly crystallize in space .
other experiments were performed to study the effects of gravity on colloidal dispersions .
bailey _ _ et al.__@xcite used small - angle light scattering and direct imaging to study the phase separation in a deeply quenched colloid - polymer mixture in microgravity on the international space station ( iss ) .
the authors followed this process for nearly 5 decades in time , from the early stage of spinodal decomposition , which is governed by diffusion - limited dynamics , to the late stage governed by interfacial - tension - driven coarsening , where the domains reached a centimeter in size .
the phase separation dynamics closely mimic that of molecular fluid mixtures after very shallow quenches through the critical point .
et al.__@xcite studied the aggregation kinetics of polystyrene colloids in space and compared the results with those measured on earth . in microgravity
the clusters grew at a rate in qualitative agreement with that expected for diffusion - limited cluster aggregation ( dlca ) , which persisted for the full duration of the space experiment .
by contrast , the time evolution of the average cluster sizes on earth showed a significant deviation from dlca behavior .
the initial growth of the clusters was identical to that in space , following a power - law time evolution consistent with dlca ; however , after a certain time , the growth deviated from the expected behavior and reached a plateau , with no further growth observed .
et al.__@xcite studied the kinetics of colloidal alloy crystallization in binary mixtures of polystyrene and silica particles having different sizes and densities in microgravity using parabolic flights in an airplane .
the growth rates increased substantially in microgravity up to about 1.7 times those in normal gravity , which was mainly caused by the absence of particle segregation that took place under normal gravity .
_ et al.__@xcite studied the effect of added salt on the crystallization kinetics of silica particles .
the authors found that the rate coefficients decreased as the salt concentration increased and that those in microgravity were smaller than those in normal gravity .
_ et al.__@xcite used atomic force microscopy to study the effect of gravity on the deposition of polystyrene sulfate latex , silica and colloidal gold particles onto solid surfaces .
the results showed that significant differences are observed in the particle deposition onto horizontal and vertical surfaces , under identical suspension conditions and exposure times .
when the density of the colloidal particles is high , as in the case of colloidal gold particles , gravity can become a significant driving force for particle transport to the surface .
as mentioned , only a limited number of experiments can be and have been performed in true microgravity conditions .
it is therefore important to diagnose whether gravity may significantly influence the outcome of a ground - based experiment .
bartlett _ _
et al.__@xcite proposed a simple and elegant way of simulating time - averaged _ zero_-gravity conditions on earth .
the method consists simply of rotating the samples slowly ( with ca . 1 hz ) in the vertical plane at a certain angular velocity , @xmath26 , so the sedimentation of a particle is effectively averaged out over one cycle .
the optimum angular frequency is determined by a balance between the rate of sedimentation and the magnitude of the rotationally induced shear field within a sample .
time scales of other dynamical processes such as phase separation should be taken into account as well .
in fact , @xmath26 must be sufficiently large to ensure that composition gradients at any instant of time are not significant over the diameter of for instance an aggregate .
furthermore , @xmath26 must also be sufficiently small to ensure that the rotationally induced shear gradient and/or convection does not significantly distort the equilibrium microstructure . with our experiments we roughly aimed at reducing settling of particles within one rotation to the particle size or less . with this technique , the authors studied binary mixtures of large and small spherical particles , and compared the results with those subjected to normal gravity .
the overall volume fraction of the suspension , @xmath27 was 56.6% with a mixing ratio of ab@xmath28 . under normal gravity , in the top 1 mm of the system an irregularly stacked close - packed crystal of small spheres ( species b ) formed together with an ordered ab@xmath29 alloy structure .
lower down in the sample , the authors observed 15 mm of a single crystalline phase of small spheres , with the large spheres in amorphous structure . at the bottom , the system formed 10 mm of an amorphous structure .
the observation of ab@xmath29 confirmed the importance of gravity in these binary mixtures since equilibrium studies had shown that ab@xmath29 structures are formed only in suspensions with compositions between ab@xmath30 and ab@xmath31.@xcite by contrast , after slowly rotating the sample for 138 days and standing ( @xmath32 ) for 40 hours , the dispersions remained homogeneous and crystallization proceeded throughout the sample with a crystalline phase of small spheres with the remaining , relatively small number of , large spheres in amorphous structures .
amazingly , the methodology of bartlett _
et al_. has not seen that much follow - up work , although it was reinvented at least once in work on the effects of depletion on micron - sized colloidal platelets by badaire _
et al_.@xcite in this paper , we will , mostly qualitatively , study several systems that we feel are representative of many self - assembling systems being studied by many groups .
we investigate the difference between systems left to evolve under the influence of gravity and the same systems for which a slow continuous rotation averaged out the effect of particle sedimentation ( see fig .[rotstage ] ) . ultimately , it is our goal to find out under which conditions the methodology of averaging out gravity will give similar results as true _ zero_-gravity experiments and under which conditions it will not
. although it is not part of the present paper , having a confocal scan head attached to the rotation stage is in principle possible , and may provide single particle level trajectories of the systems that in this paper are studied more qualitatively .
the systems we studied include : a hard sphere fluid , a colloidal crystal of particles interacting via long - range repulsions , an oppositely charged colloidal gel that is crystallizing , a colloidal system with short - range depletion attraction and long - range electrostatic repulsion , dipolar colloidal chains , and colloidal gold platelets with attractive van der waals interactions .
to do so , we prepared identical pairs of samples of these different systems . for each case , the sample was rigorously homogenized on a vortex apparatus and transferred into two glass capillaries .
subsequently , one sample was stored horizontally under normal gravity ( see fig . [ rotstage].c , except in the case of charged colloidal crystal ( ccc ) where the sample was stored verticlally and observed with a tilted confocal microscope ) , whereas the other one was rotated in a home - built rotating stage to average out the effect of particle sedimentation ( see fig .
[ rotstage].a , b , d ) . after a number of days
, we compared the two samples by studying them with confocal microscopy .
[ cols= " < ,
< , < , < , < " , ] to average out the effects of gravitational sedimentation , we built a rotating stage using the design by badaire _ _ et al.__@xcite ( see fig .
[ rotstage].a , b ) that enabled us to rotate samples at rates between 0 and 10 rpm .
this design consists of a sample holder , made from a silicone elastomer , in which several slots were molded to accommodate microscope glass slides , and a peristaltic pump installed in such a way that the sample holder attached to its rotor rotates in a vertical plane .
the fabrication of the stage required 4 main steps : 1 ) a mold was fabricated from a rigid assembly of 28 vertically and uniformly distributed glass slides ( 26 mm long , 1 mm thick ) inserted into the lid of a 145 mm plastic petri dish .
the vertical position of the slides was adjusted such that a gap of approximately 1 mm remained at the bottom of the petri dish after positioning the lid over it .
2 ) a silicone elastomer was prepared by mixing 100 ml of sylgard dc 184 elastomer base and 10 ml of curing agent .
once degassed in a dessicator connected to a vacuum pump , the elastomer was poured into the petri dish , and the lid with the glass slide assembly was placed over it .
3 ) the whole assembly was placed in an oven at 80@xmath20c for 24 h to cure the elastomer .
4 ) once at ambient temperature , the sample holder was separated from the mold with the help of a few drops of ethanol , and screwed to the rotor of an inexpensive peristaltic pump ( variable - flow peristaltic pump , fisher scientific ) , installed on a support in such a way that the sample holder rotates exactly in a vertical plane parallel to gravity .
we also constructed a setup in which it was possible to rotate the samples while simultaneously putting them under the influence of a well defined electric field .
the setup was build from an inexpensive motor and gear assembly and a mercury wetted rotating contact ( mercotac model 305 ) .
the mercury in this `` slip ring '' ensures there is absolutely no contact bounce present in the signal . with this setup an electric signal of 250v with a frequency ranging from dc to 1 mhz
could be applied to the rotating sample .
since the samples in the stage rotate slowly , sedimentation is effectively averaged out and additionally , convection is reduced . however , since the confocal microscope setup is presently not designed for rotation , particle sedimentation could not be averaged out during the observation of the samples .
we believe that the observed structures were not affected by gravity during the short time ( @xmath33 min ) between taking the samples out of the rotation stage and the 3d - imaging .
the different dispersions studied in this work were transferred each into two glass capillaries of @xmath34 mm inner dimensions and approximately @xmath35 cm long ( vitrocom , see fig .
[ rotstage].c ) .
the samples were stored horizontally under the influence of gravity ( except in the case of charged colloidal crystal ( ccc ) where the sample was stored verticlally ) and on the rotating stage for several days .
the capillaries were treated in an identical way as well and were closed with the same glue ( norland optical adhesive ) , which was cured for the same amount of time with uv light ( @xmath36 nm , uvgl-58 uv lamp , uvp ) .
care was taken to avoid inclusion of an air bubble in the sample , which could disrupt colloidal structures when on the rotating stage . with the confocal microscope
, we observed that in both capillaries for each system the same state was present initially .
table [ tablepmmas ] gives some details about the different systems used in this study , labeled hsf for the hard sphere fluid , ccc for charged colloidal crystal , occg for oppositely charged colloidal gel , cda for colloids with depletion attraction , dcc for dipolar colloidal chains and cgp for colloidal gold platelets .
_ hard - sphere _ colloidal systems are one of the simplest models used to study a variety of fundamental questions .
the phase diagram of these systems is well known , @xcite showing a fluid phase at a volume fraction less than @xmath37 , a fluid / crystal coexistence at @xmath22 , and between @xmath38 a close packed crystal .
when crystallization is avoided by introducing particle size polydispersity , hard sphere suspensions become very viscous and eventually form an amorphous solid at large volume fraction .
@xcite ) made of pmma particles , suspended in cyclohexyl bromide ( chb ) saturated with 260 @xmath15 m salt tetrabutyl ammonium bromide ( tbab).,scaledwidth=31.0% ] in this section , we investigate the effects of gravity on a dilute fluid made of fluorescently labelled and sterically stabilized poly methyl - methacrylate ( pmma ) particles with a diameter of 2.2 @xmath15 m and size polydispersity of @xmath39 , dispersed in cyclohexyl bromide ( no density matching , @xmath40 g / ml ) saturated with added salt tetrabutyl ammonium bromide ( 260 @xmath15 m tbab , sigma - aldrich ) .
note that the particles sediment upwards ( creaming ) since they have a lower density than the solvent . in these experiments on sterically
stabilized and refractive index matched colloids,@xcite the interparticle interactions can easily be tuned from long - range electrostatic repulsive to ( almost ) hard sphere - like , by adding salt that allows to effectively screen the colloidal surface charge.@xcite the suspension was prepared at a volume fraction @xmath41 , and was transfered into two glass capillaries of @xmath34 mm inner dimensions ( vitrocom ) .
the samples were stored horizontally in one case and on the rotating stage in the other case . with the confocal microscope
, we observed that in both capillaries the same fluid - like state ( see fig . [ fluid5percpmmachb ] ) was present initially . ) at a height @xmath42 m , ( b ) dense fluid phase ( @xmath43 ) at @xmath44 m , ( c ) fluid / crystal coexistence ( @xmath45 ) at @xmath46 m and ( d ) close packed crystal ( @xmath47 ) at @xmath48 m . all distances are measured from the top of the cell .
the inset in ( a ) shows an @xmath49 image of 38 @xmath15 m sediment , and in ( b ) , ( c ) and ( d ) show 2 times zoomed images.,scaledwidth=45.0% ] after a few tens of minutes , significant sedimentation was already observed in the sample that was stored horizontally .
two time scales are relevant : that of diffusion ; here the brownian time is defined as the time it takes a particle to diffuse over a distance of one radius , @xmath50 seconds , and that of gravity induced settling , which can be characterized by the time it takes a particle to sediment over one radius , @xmath51 seconds . as already mentioned in the introduction , the ratio of these time scales is quantified by the pclet number @xmath52 ( gravitational length @xmath53 m ) .
et al_.@xcite and schmidt _ et al_.@xcite quantitatively studied the nonequilibrium sedimentation of hard - sphere dispersions with pclet number @xmath54 , that were confined in horizontal capillaries using confocal microscopy , dynamical density functional theory ( ddft ) , and brownian dynamics simulations .
the authors showed that ddft gives accurate results as compared to computer simulations and that it describes laterally homogeneous sedimentation very well as compared to experimental data .
hoogenboom _ _ et al.__@xcite studied in real - space crystallization and the occurrence of stacking faults formed through sedimentation of ( soft ) silica colloids with pclet numbers on the order of @xmath55 .
the crystals showed a high amount of f.c.c - stacking compared to bulk crystallization of hard - spheres . in our case ,
the system has a stacking probability not close to that of f.c.c , but closer to r.h.c.p ( random hexagonal closed packed structure ) .
note that the density profile could have been measured and from this the equation of state could have been obtained.@xcite after a few hours , crystallization occured on the top of the capillary cell ( pmma particles cream in chb ) .
note that the time needed for a particle to sediment over the capillary thickness ( ca . @xmath56
m ) is on the order of a few tens of minutes . on the other hand , particles in the sample stored in the rotating stage
were homogeneously distributed and did not display any significant density gradient .
observed one week later , the particles in the sample that was stored horizontally reached an equilibrium density profile ( fig .
[ sedimentedfluid5percpmmachb ] ) .
however , no significant change was observed on the rotated sample .
we have checked that rotating the hard sphere fluid does not alter the particle distribution , by comparing the measured radial distribution functions ( not shown here ) for the sample just after preparation and after more than a week of a storage in the rotating stage .
colloidal particles , suspended in a solvent with a ( relative ) dielectric constant roughly between 4 and 15 , can form very long - range crystal lattices even for micron - sized particles when the suspension is carefully purified to remove free ions from the suspension .
@xcite a suspension of pmma particles suspended in purified cyclohexyl bromide chb ( dielectric constant @xmath57 ) is one example of such systems . in these kinds of solvents ,
charge dissociation still occurs spontaneously @xcite , contrary to truly apolar media ( @xmath58 ) that require charge stabilizing surfactants.@xcite the particles used in this study carried a positive charge , @xmath59 electrons ( as quantified by means of electrophoresis measurements of the dilute system ) , and were surrounded by a double layer of compensating negative charges .
each particle exerts a significant screened coulomb force on its neighbors , confining the brownian motion to excursions around lattice positions because the array crystallizes under the influence of purely repulsive forces .
the interparticle spacing can be as much as ten times the particle radius , as shown in fig .
[ crystal1percpmmachb ] , which is typical of _ wigner crystals_.@xcite m , suspended in pure cyclohexyl bromide ( chb ) .
the volume fraction @xmath60 and the interparticle spacing is @xmath61m.,scaledwidth=31.0% ] we observed the effect of gravitational compression on a long ranged crystal of pmma particles with diameter of @xmath62 m , suspended in pure cyclohexyl bromide ( chb ) at a volume fraction @xmath60 ( see fig .
[ crystal1percpmmachb ] ) . to reach a very low ionic strength condition , we purified the solvent chb using a procedure described elsewhere,@xcite and put the suspension in contact with water that serves as a reservoir absorbing the free ions from the suspensions.@xcite the conductivity of chb as received ( ca .
@xmath63 ps / cm ) decreased to less than @xmath64 ps / cm with the purification procedure . due to the small volume of the dispersion contained inside the capillaries
, we could not directly measure the conductivity of the oil - water combination in our experiment , but it was previously shown@xcite that upon contact with an equal volume of water the conductivity of chb decreases by a factor of 10 . from this , we estimated the debye screening length to be @xmath65 m in this case . the crystal structure
observed at this low particle volume fraction and for our experimental condition of the screening @xmath66 has a b.c.c symmetry , as b.c.c under these conditions has a lower free energy state than f.c.c.@xcite the capillary cell was placed vertically in this case and observed with a 90@xmath20 tilted confocal scanhead combined with an accurate translation stage that allows to obtain 3d data stacks at all the different heights along the length of the capillary ( which is ca . 3 cm in this case ) .
the water phase was located above the chb phase , since water has a lower density than chb . ) at a height @xmath67 mm below the water interface .
long range crystals : ( b ) a ( 110 ) plane of a b.c.c crystal ( @xmath68 ) at @xmath69 mm , ( c ) a ( 110 ) plane of a b.c.c crystal ( @xmath70 ) at @xmath71 mm and ( d ) a ( 111 ) plane of a f.c.c crystal ( @xmath72 ) at @xmath73 mm . the lattice spacing varies in the crystal phase from more than 10 @xmath15 m to 3 - 4 @xmath15 m .
( e ) a reentrant dense fluid ( @xmath74 ) at @xmath75 mm ( particles diffuse freely ) and ( f ) a ( 111 ) plane of a f.c.c crystal ( @xmath76).,scaledwidth=46.0% ] figure [ sequence ] shows the nonequilibrium situation of the system after it was stored vertically for 1 day under the influence of gravity . the system formed a dilute fluid ( in the lowest region of the capillary because the particles have a lower density than the solvent ) , followed by long - range crystals with different lattice spacings , varying from more than 10 @xmath15 m to 3 - 4 @xmath15 m .
going further upwards , we observed a reentrant dense fluid phase until finally the system crystallized again at a much higher volume fraction . at this low ionic strength condition and increasing the particle volume fraction by gravity , the system showed a transition from a fluid phase , to low density crystals of b.c.c structure , to f.c.c crystals , as confirmed previously by experiments and numerical simulations.@xcite a dramatically different phase behavior was observed in the other ( initially identical ) capillary that was left in the rotating stage . during all this time and throughout the whole capillary the low volume fraction crystal as depicted in fig .
[ crystal1percpmmachb ] was preserved .
the unusual sequence dilute - fluid / dilute - crystal / dense - fluid / dense - crystal phases observed throughout the suspension under the influence of gravity is likely a consequence of the volume fraction dependence of the interaction potential , an observation made before on a similar system by some of the authors.@xcite in more recent work , we measured interaction forces between pairs of charged pmma colloidal particles suspended in an equivalent solvent ( cyclohexyl chloride , chc ) directly from the deviations of particle positions inside two time - shared optical traps.@xcite in addition , from electrophoresis experiments it was observed that the particle charge in similar systems as the one used here decreases non - linearly with increasing volume fraction.@xcite current work is in progress using this method to probe the interaction forces in this system more directly with optical tweezers in more concentrated situations as observed in this section , which will allow for a better understanding of several of the open questions related to , for instance , the presence of non - additive many - body interactions in these long - ranged charged systems.@xcite after more than 3 days , the observed sequence of reentrant crystal and liquid phases collapsed into a less than 2 mm thick layer of dense crystals with few layers of a fluid phase just below .
we note again that the dilute crystal on the rotating stage remained long ranged for more than a week of observation . as mentioned above , we believe that this intriguing behavior observed in this system while it is in transition to its equilibrium state in the gravitational field is caused mainly by a density - dependent surface charge in our system .
particles lose their charges during densification but not completely , which explains why the system does not reach the hard sphere crystal . at this point
we conjecture that this effect of colloidal discharging@xcite is responsible for the huge variation of the lattice spacing in the sedimented crystal and for the observed reentrant fluid phase .
indeed , densification of the system under gravity and , consequently , the colloidal discharging make the competition between electrostatic interaction energy and thermal energy quite complex.@xcite these effects are the subject of work in progress .
since a few years , binary systems of oppositely charged colloids are used to create so - called _ ionic colloidal crystals _ in low polar solvent mixtures.@xcite particles can be made oppositely charged by tuning the salt concentration and the volume fractions of both particle species.@xcite when the attractions are strong ( @xmath77 ) , the oppositely charged particles strongly attract each other and become kinetically trapped in a gel - like structure.@xcite when the charges are sufficiently low and the interactions are on the order of @xmath78 , the system can efficiently explore phase space , resulting in the formation of oppositely charged binary crystals . the phase diagram for such systems
was calculated , and cscl and cuau and random f.c.c structures were found both in computer simulations and experiments.@xcite @xmath15 m tbab , to a capillary cell .
the total volume fraction @xmath79.,scaledwidth=33.0% ] to study the gravity induced collapse of a gel - like structure made of oppositely charged colloids , we prepared a 1:1 mixture ( in volume ) of pmma particles with diameters of 1.06 and 0.91 @xmath15 m ( see table [ tablepmmas ] ) in a nearly , but not perfectly , density matched mixture of chb / cis - decalin ( 27.2 w% ) containing tetrabutyl ammonium bromide salt ( 26 @xmath15 m tbab , sigma - aldrich ) ( see fig .
[ ch5:day0small ] ) .
the estimated debye screening length is @xmath80 m and the suspension was prepared at a total volume fraction @xmath79 .
days of horizontal storage ( a , b ) or in the rotating stage ( c , d ) . due to sedimentation ,
the gel - like structure had partially collapsed and cscl crystal domains ( ca .
@xmath15 m in size ) were present on top of the sediment ( grey line in ( b ) ) .
the distance from the lower glass wall to the top of the sedimented part ( including the crystals ) was @xmath82 @xmath15 m .
for the sample stored in the rotating stage , the gel - structure survived and small crystal domains ( ca .
@xmath15 m in size ) had emerged around the voids inside it .
the thick grey horizontal lines in ( b , d ) denote the position of the capillary glass walls . image ( b ) and ( d ) are composite images that are build up from two or three confocal images.,scaledwidth=46.0% ] after @xmath84 days of horizontal storage , the gel had partially collapsed and the particles were visibly sedimented ( fig .
[ ch5:day18small]b ) .
crystal domains of only several layers thick were present on top of the sediment ( fig .
[ ch5:day18small]a ) . for the sample that was stored on the rotating stage ,
the gel - structure was still present and no signs of sedimentation were observed ( fig .
[ ch5:day18small]d ) .
the gel had contracted and pulled away from both walls due to the cohesive forces .
interestingly , crystal formation occurred at the interface between the gel and the voids inside the gel ( fig .
[ ch5:day18small]c ) .
days of horizontal storage ( a , b ) or in the rotating stage ( c , d ) . under the influence of gravity ,
the crystals on top of the sediments that were already present after @xmath84 days had grown longer in the direction parallel to the gravitational field .
the distance from the lower glass wall to the top of the crystals was ca .
@xmath15 m . on the rotating stage ,
the gel - like structure had completely disappeared and all particles were present in crystal domains .
there were large voids between different crystal domains .
the grey horizontal lines in ( b , d ) denote the positions of the capillary glass walls .
image ( b ) and ( d ) are composite images that are build up from two or three confocal images.,scaledwidth=47.0% ] after spending a total time of @xmath85 days of horizontal storage , the crystals that were present at the top of the sediment had grown out in an elongated fashion ( fig .
[ ch5:day30small]b ) .
since the distance from the top of the crystallites to the lower glass wall of the sample did not change much compared to the situation after @xmath84 days ( fig . [ ch5:day18small]a , b ) , the crystal growth direction must have been downwards into the collapsed gel .
apparently , the voids in the sediment offered sufficient translational freedom for particle reorganization that was required for the conversion of the sediment into the crystal .
approximately @xmath86 @xmath15 m above the lower glass wall , the growth process stalled , suggesting that the sediment became too dense to allow for the particles to rearrange .
contrary to this , for the sample that was stored for a total time of @xmath85 days on the rotating stage , the gel - like structure had completely disappeared and crystallites had appeared with random orientations throughout the whole sample ( fig . [ ch5:day30small]c , d ) .
the orientation of the crystal domains appeared random , indicating that the small crystals that were already present after @xmath84 days had continued to grow until all of the gel had converted into crystal .
the large voids that were present between different crystals show that colloidal sedimentation was efficiently averaged out .
m in chb / cis - decalin with @xmath87 @xmath15 m tbab after vigorous shaking ( @xmath88 ) .
( b , c ) dispersions after @xmath89 hours of horizontal storage show that the particles sedimented . on top of the sediment , some very small crystalline domains were visible .
( d - f ) after @xmath89 hours , the dispersion on the rotating stage contained randomly oriented crystal domains with large voids in between , similar to fig .
[ ch5:day30small ] .
the thick grey horizontal lines in ( c , f ) denote the positions of the capillary glass walls . the upper glass wall in ( c ) is not shown here.,scaledwidth=46.5% ] additionally , we examined crystal growth starting from a fluid of oppositely charged particles ( with diameter of @xmath90 m in this case ) at a total volume fraction , @xmath88 , ( see fig .
[ ch5:day0big]a ) . under the influence of gravity the particles sedimented within @xmath89 hours , reducing the region containing particles to approximately @xmath91 @xmath15 m ( fig . [ ch5:day0big]c ) . on top of the sediment , small crystals were sporadically present ( fig .
[ ch5:day0big]b ) .
the identical dispersion that had spent the same time on the rotating stage , however , had crystallized completely ( fig .
[ ch5:day0big]d - f ) .
similar to the gel - samples , gravity and the local volume fraction play , as expected , an important role in the crystal growth of micron - sized colloids .
our results show that for these systems of oppositely charged particles , despite being nearly density matched , densification by sedimentation makes crystal growth much slower and in some cases inhibits it almost completely .
contrary to the previous system of oppositely charged colloids , attraction between colloidal particles can also be induced by adding non - adsorbing polymers to the dispersion.@xcite this polymer - mediated _ depletion interaction _ has received significant attention in the last decades . by varying the polymer concentration and the size ratio between colloids and polymers , one can control the strength and the range of the interparticle interaction potential , respectively . here
we study the situation where depletion interactions are supplemented by and compete with a long - range electrostatic repulsion . at low colloid concentrations , experiments , computer simulations and theory
have reported a stable colloidal cluster phase.@xcite . and
the attraction strength @xmath92.,scaledwidth=31.0% ] in this section , we have verified qualitatively the effects of gravity on the cluster size distribution of a dispersion with competing short - range attraction and long - range coulomb repulsion . in these systems , the average cluster size @xmath93 grows with the overall volume fraction of the colloids @xmath94 as @xmath95 as predicted by theory@xcite and confirmed by experiments@xcite .
sedimentation in a gravitational field couples to the local colloid volume fraction , which in turn leads to larger colloidal clusters .
ultimately these clusters aggregate amongst themselves , forming a colloidal gel .
the system studied consists of pmma - particles dispersed in a mixture of cyclohexyl bromide ( chb ) and cis - decalin with a short - range depletion attraction that is induced by adding non - adsorbing polystyrene polymers .
more details of this system can be found in a recent paper by zhang _
et al_.@xcite the volume fraction of the dispersion we studied was @xmath41 and the attraction strength was @xmath92 .
the system formed a cluster phase with clusters containing each few tens of particles on average , as shown in fig .
[ geldepletion ] .
we prepared two samples , one left horizontally and the other one stored in the rotating stage .
after a few hours of sedimentation , the clusters had grown and became gradually interconnected , forming a gel with increasing local volume fraction ( see fig .
[ depletion]b ) . on the other hand
, the rotated sample preserved the same cluster phase during this time as depicted in fig .
[ depletion]c .
rotational averaging out gravitational sedimentation avoids gel formation in systems with short - range attraction and long - range repulsion . instead
, the cluster phase is stable .
clearly the cluster phase was not destabilized ( shear - melted ) by rotation of the sample .
m and @xmath96 m from the bottom of the cell , respectively .
( c ) rotated sample .
( d ) and ( e ) are @xmath49 images of the horizontally stored sample and the rotated one respectively .
note the absense of a gel phase in the rotated sample.,scaledwidth=46.5% ] colloidal particles whose dielectric constant is different from that of the solvent , acquire a dipole moment parallel to an external electric field .
the induced _ dipolar interactions _ between the particles are anisotropic and long - ranged .
the dipolar strength depends on the local particle concentration , particle size , the dielectric constant mismatch , and the applied electric field strength.@xcite the structure of the dispersion can be tuned from strings ( 1d ) , to sheets ( 2d ) , and eventually to several different types of equilibrium 3d crystal structures by varying the dipolar field strength and concentration . in the low - field regime ,
the stable structure found in both experiments , theory and computer simulations is a string fluid phase that consists of chains of particles parallel to the field direction and with liquid like order perpendicular to the field direction.@xcite here , we studied the effect of gravity on the string length distribution with a system of electrostatically stabilized silica particles ( @xmath97 , diameter @xmath98 @xmath15 m ) suspended in dimethyl sulfoxide ( @xmath99 nm and @xmath100 ) and subjected to an ac electric field .
prior to the experiment , two identical electric cells were filled with the suspension and then rotated slowly in a plane parallel to gravity to maintain the particles homogeneously dispersed throughout the cell .
after 3 hours of rotation , one cell was removed from the rotating stage and then stored horizontally while the other one was left in the rotating stage .
subsequently , both cells were exposed to an external electric field . here
, we used another rotating stage that especially was designed and build to accommodate an electric field while the sample was rotating .
moreover , the electric field was supplied through a mercury contact to provide an uninterrupted connection , as explained in sec .
[ raogssection ] . upon application of an external electric field
( @xmath101 v/@xmath15 m , @xmath102 = 1 mhz the root - mean - square electric field and the frequency respectively ) , the induced dipole moment in each particle led the particles to assemble into strings of one particle thick that were aligned in the field direction in a head - to - tail arrangement with a broad distribution in the length of the chains.@xcite these strings then sedimented when the cell was kept horizontal , due to the density mismatch between the particles ( 2.1 g / ml ) and the solvent ( 1.1 g / ml ) . at the bottom of the cell the local concentration of the strings
slowly increased .
both ends of the longer strings continuously caught particles or small strings moving around them , thus growing into longer strings ( fig.[strings]b ) .
the distance between the individual strings was about 20 - 35@xmath4 .
therefore , the dipolar contribution between these strings was unimportant and we did not observe a further change in the string length distribution . on the other hand , a distribution of short strings ( fig.[strings]a ) was maintained homogeneously throughout the dispersion by means of rotating the cell .
our experimental results qualitatively demonstrate that the structure evolution , e.g. string length distribution , becomes quite different when the sedimentation of colloids is averaged out by rotating the sample .
when dispersed in water , highly anisotropic single crystalline gold platelets ( @xmath103 m diameter , @xmath104 nm thick ) can self - assemble into long wormlike columns by finely tuning the interplay between van der waals and electrostatic interactions.@xcite in absence of salt , and in the limit of low volume fraction ( @xmath105 ) , the particles remain isotropically dispersed , but for ionic strengths comprised roughly between 0.8 and 2 mm , the particles form dynamic columns or stacks , the length of which depends on the depth of the secondary van der waals minimum in the pair potential and on the volume fraction .
however , because of the high density mismatch between gold and water , the particles sediment quickly in a stationary capillary and the observed structures are therefore highly influenced by gravity . since , in this system , the main consequence of gravity is an increase of the local particle concentration near the surface , the length of the columns ( at constant ionic strength ) is expected to be longer after the capillary is allowed to stay horizontal for a sufficient amount of time .
we verified this experimentally by comparing qualitatively the column size distribution in a suspension in which the columns were kept homogeneously dispersed throughout the whole volume with the help of the rotating stage and in an identical suspension that was kept horizontal . the first image ( a ) of fig . [ gold
] was captured after allowing the capillary to rotate on the stage for 72 hours after preparation ( the image was taken less than 1 min after removing the capillary from the stage ) .
the ionic strength in the suspension was 1.5 mm ( nacl ) .
only short columns were seen .
a gas phase of very short columns and individual particles was present in the whole volume above this first layer of columns because of the very large difference in the gravitational length between individual particles and the stacks of particles .
the second image ( b ) was captured after keeping the same capillary immobile for 24 hours .
it can be clearly seen that the size distribution evolved to much larger lengths , and few short columns or isolated particles were observed .
it was checked qualitatively that the same size distribution was obtained by preparing another capillary and leaving it undisturbed for 72 hours before observation .
in this paper , we studied the effect of particle sedimentation on the structure formation of a whole range of different self - assembling colloidal systems , including a hard sphere fluid , a dispersion with a long debye screening length at a density above the freezing transition , an oppositely charged colloidal gel that is crystallizing , colloids with competing short - rang depletion attraction and long - rang electrostatic repulsion , colloidal dipolar chains , and colloidal gold platelets under conditions in which they organized into equilibrium stacks .
we used a simple home - built rotating stage to average out the effect of particle sedimentation , and compared the results with identical samples that were left under the influence of gravity for the same amount of time .
in addition , we showed how such a setup can also be extended and used in cases were mhz frequency externally applied electric fields are used to manipulate a dispersion . in the case of the repulsive charged colloidal crystal , an unusual transient sequence of dilute - fluid /
dilute - crystal / dense - fluid / dense - crystal phases was observed throughout the suspension under the influence of gravity , which is related to the volume fraction dependence of the interactions between charged particles . in the case of oppositely charged colloids ,
a gel - like structure was found to collapse under the effect of gravity with a few crystalline layers formed on top of the sediment , whereas the gel completely transformed into crystallites with random orientations when sedimentation was averaged out by slowly rotating the sample .
we showed in general that the use of a rotating stage is very useful to diagnose when gravity may significantly influence the outcome of ground - based experiments .
moreover , we demonstrated that it can serve as a cheap and simple method to average out the effects of particle sedimentation for a wide variety of colloidal systems .
further work needs to find the conditions under which this method approaches systems at _ zero_-gravity preferably by comparison with similar experiments under true ( micro ) zero - gravity conditions .
it will also be worthwhile to be able to follow structuring quantitatively in real space while the rotations are actually taking place .
this work was supported by the research program _ colloids in space _ , financed by nwo - sron .
tv and js acknowledge nwo - cw for financial support . s.b . and h.r.v .
were supported by the research program of the stichting voor fundamenteel onderzoek der materie , financed by the nederlandse organisatie voor wetenschappelijk onderzoek ( nwo ) .
part of this research was supported by the eu ( nanodirect , grant number cp - fp-213948 - 2 ) .
bailey , w.c.k .
poon , r.j .
christianson , a.b .
schofield , u. gasser , v. prasad , s. manley , p.n .
segre , l. cipelletti , w.v .
meyer , m.p .
doherty , s. sankaran , a.l .
jankovsky , w.l .
shiley , j.p .
bowen , j.c .
eggers , c. kurta , t. lorik jr .
pusey , and d.a .
weitz , phys .
lett . , 2007 , 99 , 205701 .
s. manley , l. cipelletti , v. trappe , a.e .
bailey , r.j .
christianson , u. gasser , v. prasad , p.n .
segre , m.p .
doherty , s. sankaran , a.l .
jankovsky , b. shiley , j. bowen , j. eggers , c. kurta , t. lorik , and d.a .
weitz , phys .
lett . , 2004 , 93 , 108302 . | we report on the differences between colloidal systems left to evolve in the earth s gravitational field and the same systems for which a slow continuous rotation averaged out the effects of particle sedimentation on a distance scale small compared to the particle size .
several systems of micron - sized colloidal particles were studied : a hard sphere fluid , colloids interacting via long - range electrostatic repulsions above the freezing volume fraction , an oppositely charged colloidal system close to either gelation and/or crystallization , colloids with a competing short - range depletion attraction and a long - range electrostatic repulsion , colloidal dipolar chains , and colloidal gold platelets under conditions where they formed stacks .
important differences in the structure formation were observed between the experiments where the particles were allowed to sediment and those where sedimentation was averaged out .
for instance , in the case of colloids interacting via long - range electrostatic repulsions , an unusual sequence of dilute - fluid / dilute - crystal / dense - fluid / dense - crystal phases was observed throughout the suspension under the effect of gravity , related to the volume fraction dependence of the colloidal interactions , whereas the system stayed homogeneously crystallized with rotation . for the oppositely charged colloids ,
a gel - like structure was found to collapse under the influence of gravity with a few crystalline layers grown on top of the sediment , whereas when the colloidal sedimentation was averaged out , the gel completely transformed into crystallites that were oriented randomly throughout the sample .
rotational averaging out gravitational sedimentation is an effective and cheap way to estimate the importance of gravity for colloidal self - assembly processes . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.